langchain/docs/docs/additional_resources/arxiv_references.mdx

# arXiv
            
LangChain implements the latest research in the field of Natural Language Processing.
This page contains `arXiv` papers referenced in the LangChain Documentation, API Reference,
and Templates.

## Summary

| arXiv id / Title | Authors | Published date 🔻 | LangChain Documentation|
|------------------|---------|-------------------|------------------------|
| `2312.06648v2` [Dense X Retrieval: What Retrieval Granularity Should We Use?](http://arxiv.org/abs/2312.06648v2) | Tong Chen, Hongwei Wang, Sihao Chen,  et al. | 2023-12-11 | `Template:` [propositional-retrieval](https://python.langchain.com/docs/templates/propositional-retrieval)
| `2311.09210v1` [Chain-of-Note: Enhancing Robustness in Retrieval-Augmented Language Models](http://arxiv.org/abs/2311.09210v1) | Wenhao Yu, Hongming Zhang, Xiaoman Pan,  et al. | 2023-11-15 | `Template:` [chain-of-note-wiki](https://python.langchain.com/docs/templates/chain-of-note-wiki)
| `2310.06117v2` [Take a Step Back: Evoking Reasoning via Abstraction in Large Language Models](http://arxiv.org/abs/2310.06117v2) | Huaixiu Steven Zheng, Swaroop Mishra, Xinyun Chen,  et al. | 2023-10-09 | `Template:` [stepback-qa-prompting](https://python.langchain.com/docs/templates/stepback-qa-prompting)
| `2305.14283v3` [Query Rewriting for Retrieval-Augmented Large Language Models](http://arxiv.org/abs/2305.14283v3) | Xinbei Ma, Yeyun Gong, Pengcheng He,  et al. | 2023-05-23 | `Template:` [rewrite-retrieve-read](https://python.langchain.com/docs/templates/rewrite-retrieve-read)
| `2305.08291v1` [Large Language Model Guided Tree-of-Thought](http://arxiv.org/abs/2305.08291v1) | Jieyi Long | 2023-05-15 | `API:` [langchain_experimental.tot](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.tot)
| `2303.17580v4` [HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face](http://arxiv.org/abs/2303.17580v4) | Yongliang Shen, Kaitao Song, Xu Tan,  et al. | 2023-03-30 | `API:` [langchain_experimental.autonomous_agents](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.autonomous_agents)
| `2303.08774v6` [GPT-4 Technical Report](http://arxiv.org/abs/2303.08774v6) | OpenAI, Josh Achiam, Steven Adler,  et al. | 2023-03-15 | `Docs:` [docs/integrations/vectorstores/mongodb_atlas](https://python.langchain.com/docs/integrations/vectorstores/mongodb_atlas)
| `2301.10226v4` [A Watermark for Large Language Models](http://arxiv.org/abs/2301.10226v4) | John Kirchenbauer, Jonas Geiping, Yuxin Wen,  et al. | 2023-01-24 | `API:` [langchain_community.llms...OCIModelDeploymentTGI](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.oci_data_science_model_deployment_endpoint.OCIModelDeploymentTGI.html#langchain_community.llms.oci_data_science_model_deployment_endpoint.OCIModelDeploymentTGI), [langchain_community.llms...HuggingFaceTextGenInference](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference.html#langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference), [langchain_community.llms...HuggingFaceEndpoint](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint.html#langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint)
| `2212.10496v1` [Precise Zero-Shot Dense Retrieval without Relevance Labels](http://arxiv.org/abs/2212.10496v1) | Luyu Gao, Xueguang Ma, Jimmy Lin,  et al. | 2022-12-20 | `API:` [langchain.chains...HypotheticalDocumentEmbedder](https://api.python.langchain.com/en/latest/chains/langchain.chains.hyde.base.HypotheticalDocumentEmbedder.html#langchain.chains.hyde.base.HypotheticalDocumentEmbedder), `Template:` [hyde](https://python.langchain.com/docs/templates/hyde)
| `2212.07425v3` [Robust and Explainable Identification of Logical Fallacies in Natural Language Arguments](http://arxiv.org/abs/2212.07425v3) | Zhivar Sourati, Vishnu Priya Prasanna Venkatesh, Darshan Deshpande,  et al. | 2022-12-12 | `API:` [langchain_experimental.fallacy_removal](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.fallacy_removal)
| `2211.13892v2` [Complementary Explanations for Effective In-Context Learning](http://arxiv.org/abs/2211.13892v2) | Xi Ye, Srinivasan Iyer, Asli Celikyilmaz,  et al. | 2022-11-25 | `API:` [langchain_core.example_selectors...MaxMarginalRelevanceExampleSelector](https://api.python.langchain.com/en/latest/example_selectors/langchain_core.example_selectors.semantic_similarity.MaxMarginalRelevanceExampleSelector.html#langchain_core.example_selectors.semantic_similarity.MaxMarginalRelevanceExampleSelector)
| `2211.10435v2` [PAL: Program-aided Language Models](http://arxiv.org/abs/2211.10435v2) | Luyu Gao, Aman Madaan, Shuyan Zhou,  et al. | 2022-11-18 | `API:` [langchain_experimental.pal_chain...PALChain](https://api.python.langchain.com/en/latest/pal_chain/langchain_experimental.pal_chain.base.PALChain.html#langchain_experimental.pal_chain.base.PALChain), [langchain_experimental.pal_chain](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.pal_chain)
| `2209.10785v2` [Deep Lake: a Lakehouse for Deep Learning](http://arxiv.org/abs/2209.10785v2) | Sasun Hambardzumyan, Abhinav Tuli, Levon Ghukasyan,  et al. | 2022-09-22 | `Docs:` [docs/integrations/providers/activeloop_deeplake](https://python.langchain.com/docs/integrations/providers/activeloop_deeplake)
| `2205.12654v1` [Bitext Mining Using Distilled Sentence Representations for Low-Resource Languages](http://arxiv.org/abs/2205.12654v1) | Kevin Heffernan, Onur Çelebi, Holger Schwenk | 2022-05-25 | `API:` [langchain_community.embeddings...LaserEmbeddings](https://api.python.langchain.com/en/latest/embeddings/langchain_community.embeddings.laser.LaserEmbeddings.html#langchain_community.embeddings.laser.LaserEmbeddings)
| `2204.00498v1` [Evaluating the Text-to-SQL Capabilities of Large Language Models](http://arxiv.org/abs/2204.00498v1) | Nitarshan Rajkumar, Raymond Li, Dzmitry Bahdanau | 2022-03-15 | `API:` [langchain_community.utilities...SQLDatabase](https://api.python.langchain.com/en/latest/utilities/langchain_community.utilities.sql_database.SQLDatabase.html#langchain_community.utilities.sql_database.SQLDatabase), [langchain_community.utilities...SparkSQL](https://api.python.langchain.com/en/latest/utilities/langchain_community.utilities.spark_sql.SparkSQL.html#langchain_community.utilities.spark_sql.SparkSQL)
| `2202.00666v5` [Locally Typical Sampling](http://arxiv.org/abs/2202.00666v5) | Clara Meister, Tiago Pimentel, Gian Wiher,  et al. | 2022-02-01 | `API:` [langchain_community.llms...HuggingFaceTextGenInference](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference.html#langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference), [langchain_community.llms...HuggingFaceEndpoint](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint.html#langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint)
| `2103.00020v1` [Learning Transferable Visual Models From Natural Language Supervision](http://arxiv.org/abs/2103.00020v1) | Alec Radford, Jong Wook Kim, Chris Hallacy,  et al. | 2021-02-26 | `API:` [langchain_experimental.open_clip](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.open_clip)
| `1909.05858v2` [CTRL: A Conditional Transformer Language Model for Controllable Generation](http://arxiv.org/abs/1909.05858v2) | Nitish Shirish Keskar, Bryan McCann, Lav R. Varshney,  et al. | 2019-09-11 | `API:` [langchain_community.llms...HuggingFaceTextGenInference](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference.html#langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference), [langchain_community.llms...HuggingFaceEndpoint](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint.html#langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint)
| `1908.10084v1` [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](http://arxiv.org/abs/1908.10084v1) | Nils Reimers, Iryna Gurevych | 2019-08-27 | `Docs:` [docs/integrations/text_embedding/sentence_transformers](https://python.langchain.com/docs/integrations/text_embedding/sentence_transformers)

## Dense X Retrieval: What Retrieval Granularity Should We Use?

- **arXiv id:** 2312.06648v2
- **Title:** Dense X Retrieval: What Retrieval Granularity Should We Use?
- **Authors:** Tong Chen, Hongwei Wang, Sihao Chen,  et al.
- **Published Date:** 2023-12-11
- **URL:** http://arxiv.org/abs/2312.06648v2
- **LangChain:**

   - **Template:** [propositional-retrieval](https://python.langchain.com/docs/templates/propositional-retrieval)

**Abstract:** Dense retrieval has become a prominent method to obtain relevant context or
world knowledge in open-domain NLP tasks. When we use a learned dense retriever
on a retrieval corpus at inference time, an often-overlooked design choice is
the retrieval unit in which the corpus is indexed, e.g. document, passage, or
sentence. We discover that the retrieval unit choice significantly impacts the
performance of both retrieval and downstream tasks. Distinct from the typical
approach of using passages or sentences, we introduce a novel retrieval unit,
proposition, for dense retrieval. Propositions are defined as atomic
expressions within text, each encapsulating a distinct factoid and presented in
a concise, self-contained natural language format. We conduct an empirical
comparison of different retrieval granularity. Our results reveal that
proposition-based retrieval significantly outperforms traditional passage or
sentence-based methods in dense retrieval. Moreover, retrieval by proposition
also enhances the performance of downstream QA tasks, since the retrieved texts
are more condensed with question-relevant information, reducing the need for
lengthy input tokens and minimizing the inclusion of extraneous, irrelevant
information.
                
## Chain-of-Note: Enhancing Robustness in Retrieval-Augmented Language Models

- **arXiv id:** 2311.09210v1
- **Title:** Chain-of-Note: Enhancing Robustness in Retrieval-Augmented Language Models
- **Authors:** Wenhao Yu, Hongming Zhang, Xiaoman Pan,  et al.
- **Published Date:** 2023-11-15
- **URL:** http://arxiv.org/abs/2311.09210v1
- **LangChain:**

   - **Template:** [chain-of-note-wiki](https://python.langchain.com/docs/templates/chain-of-note-wiki)

**Abstract:** Retrieval-augmented language models (RALMs) represent a substantial
advancement in the capabilities of large language models, notably in reducing
factual hallucination by leveraging external knowledge sources. However, the
reliability of the retrieved information is not always guaranteed. The
retrieval of irrelevant data can lead to misguided responses, and potentially
causing the model to overlook its inherent knowledge, even when it possesses
adequate information to address the query. Moreover, standard RALMs often
struggle to assess whether they possess adequate knowledge, both intrinsic and
retrieved, to provide an accurate answer. In situations where knowledge is
lacking, these systems should ideally respond with "unknown" when the answer is
unattainable. In response to these challenges, we introduces Chain-of-Noting
(CoN), a novel approach aimed at improving the robustness of RALMs in facing
noisy, irrelevant documents and in handling unknown scenarios. The core idea of
CoN is to generate sequential reading notes for retrieved documents, enabling a
thorough evaluation of their relevance to the given question and integrating
this information to formulate the final answer. We employed ChatGPT to create
training data for CoN, which was subsequently trained on an LLaMa-2 7B model.
Our experiments across four open-domain QA benchmarks show that RALMs equipped
with CoN significantly outperform standard RALMs. Notably, CoN achieves an
average improvement of +7.9 in EM score given entirely noisy retrieved
documents and +10.5 in rejection rates for real-time questions that fall
outside the pre-training knowledge scope.
                
## Take a Step Back: Evoking Reasoning via Abstraction in Large Language Models

- **arXiv id:** 2310.06117v2
- **Title:** Take a Step Back: Evoking Reasoning via Abstraction in Large Language Models
- **Authors:** Huaixiu Steven Zheng, Swaroop Mishra, Xinyun Chen,  et al.
- **Published Date:** 2023-10-09
- **URL:** http://arxiv.org/abs/2310.06117v2
- **LangChain:**

   - **Template:** [stepback-qa-prompting](https://python.langchain.com/docs/templates/stepback-qa-prompting)

**Abstract:** We present Step-Back Prompting, a simple prompting technique that enables
LLMs to do abstractions to derive high-level concepts and first principles from
instances containing specific details. Using the concepts and principles to
guide reasoning, LLMs significantly improve their abilities in following a
correct reasoning path towards the solution. We conduct experiments of
Step-Back Prompting with PaLM-2L, GPT-4 and Llama2-70B models, and observe
substantial performance gains on various challenging reasoning-intensive tasks
including STEM, Knowledge QA, and Multi-Hop Reasoning. For instance, Step-Back
Prompting improves PaLM-2L performance on MMLU (Physics and Chemistry) by 7%
and 11% respectively, TimeQA by 27%, and MuSiQue by 7%.
                
## Query Rewriting for Retrieval-Augmented Large Language Models

- **arXiv id:** 2305.14283v3
- **Title:** Query Rewriting for Retrieval-Augmented Large Language Models
- **Authors:** Xinbei Ma, Yeyun Gong, Pengcheng He,  et al.
- **Published Date:** 2023-05-23
- **URL:** http://arxiv.org/abs/2305.14283v3
- **LangChain:**

   - **Template:** [rewrite-retrieve-read](https://python.langchain.com/docs/templates/rewrite-retrieve-read)

**Abstract:** Large Language Models (LLMs) play powerful, black-box readers in the
retrieve-then-read pipeline, making remarkable progress in knowledge-intensive
tasks. This work introduces a new framework, Rewrite-Retrieve-Read instead of
the previous retrieve-then-read for the retrieval-augmented LLMs from the
perspective of the query rewriting. Unlike prior studies focusing on adapting
either the retriever or the reader, our approach pays attention to the
adaptation of the search query itself, for there is inevitably a gap between
the input text and the needed knowledge in retrieval. We first prompt an LLM to
generate the query, then use a web search engine to retrieve contexts.
Furthermore, to better align the query to the frozen modules, we propose a
trainable scheme for our pipeline. A small language model is adopted as a
trainable rewriter to cater to the black-box LLM reader. The rewriter is
trained using the feedback of the LLM reader by reinforcement learning.
Evaluation is conducted on downstream tasks, open-domain QA and multiple-choice
QA. Experiments results show consistent performance improvement, indicating
that our framework is proven effective and scalable, and brings a new framework
for retrieval-augmented LLM.
                
## Large Language Model Guided Tree-of-Thought

- **arXiv id:** 2305.08291v1
- **Title:** Large Language Model Guided Tree-of-Thought
- **Authors:** Jieyi Long
- **Published Date:** 2023-05-15
- **URL:** http://arxiv.org/abs/2305.08291v1
- **LangChain:**

   - **API Reference:** [langchain_experimental.tot](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.tot)

**Abstract:** In this paper, we introduce the Tree-of-Thought (ToT) framework, a novel
approach aimed at improving the problem-solving capabilities of auto-regressive
large language models (LLMs). The ToT technique is inspired by the human mind's
approach for solving complex reasoning tasks through trial and error. In this
process, the human mind explores the solution space through a tree-like thought
process, allowing for backtracking when necessary. To implement ToT as a
software system, we augment an LLM with additional modules including a prompter
agent, a checker module, a memory module, and a ToT controller. In order to
solve a given problem, these modules engage in a multi-round conversation with
the LLM. The memory module records the conversation and state history of the
problem solving process, which allows the system to backtrack to the previous
steps of the thought-process and explore other directions from there. To verify
the effectiveness of the proposed technique, we implemented a ToT-based solver
for the Sudoku Puzzle. Experimental results show that the ToT framework can
significantly increase the success rate of Sudoku puzzle solving. Our
implementation of the ToT-based Sudoku solver is available on GitHub:
\url{https://github.com/jieyilong/tree-of-thought-puzzle-solver}.
                
## HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face

- **arXiv id:** 2303.17580v4
- **Title:** HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face
- **Authors:** Yongliang Shen, Kaitao Song, Xu Tan,  et al.
- **Published Date:** 2023-03-30
- **URL:** http://arxiv.org/abs/2303.17580v4
- **LangChain:**

   - **API Reference:** [langchain_experimental.autonomous_agents](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.autonomous_agents)

**Abstract:** Solving complicated AI tasks with different domains and modalities is a key
step toward artificial general intelligence. While there are numerous AI models
available for various domains and modalities, they cannot handle complicated AI
tasks autonomously. Considering large language models (LLMs) have exhibited
exceptional abilities in language understanding, generation, interaction, and
reasoning, we advocate that LLMs could act as a controller to manage existing
AI models to solve complicated AI tasks, with language serving as a generic
interface to empower this. Based on this philosophy, we present HuggingGPT, an
LLM-powered agent that leverages LLMs (e.g., ChatGPT) to connect various AI
models in machine learning communities (e.g., Hugging Face) to solve AI tasks.
Specifically, we use ChatGPT to conduct task planning when receiving a user
request, select models according to their function descriptions available in
Hugging Face, execute each subtask with the selected AI model, and summarize
the response according to the execution results. By leveraging the strong
language capability of ChatGPT and abundant AI models in Hugging Face,
HuggingGPT can tackle a wide range of sophisticated AI tasks spanning different
modalities and domains and achieve impressive results in language, vision,
speech, and other challenging tasks, which paves a new way towards the
realization of artificial general intelligence.
                
## GPT-4 Technical Report

- **arXiv id:** 2303.08774v6
- **Title:** GPT-4 Technical Report
- **Authors:** OpenAI, Josh Achiam, Steven Adler,  et al.
- **Published Date:** 2023-03-15
- **URL:** http://arxiv.org/abs/2303.08774v6
- **LangChain:**

   - **Documentation:** [docs/integrations/vectorstores/mongodb_atlas](https://python.langchain.com/docs/integrations/vectorstores/mongodb_atlas)

**Abstract:** We report the development of GPT-4, a large-scale, multimodal model which can
accept image and text inputs and produce text outputs. While less capable than
humans in many real-world scenarios, GPT-4 exhibits human-level performance on
various professional and academic benchmarks, including passing a simulated bar
exam with a score around the top 10% of test takers. GPT-4 is a
Transformer-based model pre-trained to predict the next token in a document.
The post-training alignment process results in improved performance on measures
of factuality and adherence to desired behavior. A core component of this
project was developing infrastructure and optimization methods that behave
predictably across a wide range of scales. This allowed us to accurately
predict some aspects of GPT-4's performance based on models trained with no
more than 1/1,000th the compute of GPT-4.
                
## A Watermark for Large Language Models

- **arXiv id:** 2301.10226v4
- **Title:** A Watermark for Large Language Models
- **Authors:** John Kirchenbauer, Jonas Geiping, Yuxin Wen,  et al.
- **Published Date:** 2023-01-24
- **URL:** http://arxiv.org/abs/2301.10226v4
- **LangChain:**

   - **API Reference:** [langchain_community.llms...OCIModelDeploymentTGI](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.oci_data_science_model_deployment_endpoint.OCIModelDeploymentTGI.html#langchain_community.llms.oci_data_science_model_deployment_endpoint.OCIModelDeploymentTGI), [langchain_community.llms...HuggingFaceTextGenInference](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference.html#langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference), [langchain_community.llms...HuggingFaceEndpoint](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint.html#langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint)

**Abstract:** Potential harms of large language models can be mitigated by watermarking
model output, i.e., embedding signals into generated text that are invisible to
humans but algorithmically detectable from a short span of tokens. We propose a
watermarking framework for proprietary language models. The watermark can be
embedded with negligible impact on text quality, and can be detected using an
efficient open-source algorithm without access to the language model API or
parameters. The watermark works by selecting a randomized set of "green" tokens
before a word is generated, and then softly promoting use of green tokens
during sampling. We propose a statistical test for detecting the watermark with
interpretable p-values, and derive an information-theoretic framework for
analyzing the sensitivity of the watermark. We test the watermark using a
multi-billion parameter model from the Open Pretrained Transformer (OPT)
family, and discuss robustness and security.
                
## Precise Zero-Shot Dense Retrieval without Relevance Labels

- **arXiv id:** 2212.10496v1
- **Title:** Precise Zero-Shot Dense Retrieval without Relevance Labels
- **Authors:** Luyu Gao, Xueguang Ma, Jimmy Lin,  et al.
- **Published Date:** 2022-12-20
- **URL:** http://arxiv.org/abs/2212.10496v1
- **LangChain:**

   - **API Reference:** [langchain.chains...HypotheticalDocumentEmbedder](https://api.python.langchain.com/en/latest/chains/langchain.chains.hyde.base.HypotheticalDocumentEmbedder.html#langchain.chains.hyde.base.HypotheticalDocumentEmbedder)
   - **Template:** [hyde](https://python.langchain.com/docs/templates/hyde)

**Abstract:** While dense retrieval has been shown effective and efficient across tasks and
languages, it remains difficult to create effective fully zero-shot dense
retrieval systems when no relevance label is available. In this paper, we
recognize the difficulty of zero-shot learning and encoding relevance. Instead,
we propose to pivot through Hypothetical Document Embeddings~(HyDE). Given a
query, HyDE first zero-shot instructs an instruction-following language model
(e.g. InstructGPT) to generate a hypothetical document. The document captures
relevance patterns but is unreal and may contain false details. Then, an
unsupervised contrastively learned encoder~(e.g. Contriever) encodes the
document into an embedding vector. This vector identifies a neighborhood in the
corpus embedding space, where similar real documents are retrieved based on
vector similarity. This second step ground the generated document to the actual
corpus, with the encoder's dense bottleneck filtering out the incorrect
details. Our experiments show that HyDE significantly outperforms the
state-of-the-art unsupervised dense retriever Contriever and shows strong
performance comparable to fine-tuned retrievers, across various tasks (e.g. web
search, QA, fact verification) and languages~(e.g. sw, ko, ja).
                
## Robust and Explainable Identification of Logical Fallacies in Natural Language Arguments

- **arXiv id:** 2212.07425v3
- **Title:** Robust and Explainable Identification of Logical Fallacies in Natural Language Arguments
- **Authors:** Zhivar Sourati, Vishnu Priya Prasanna Venkatesh, Darshan Deshpande,  et al.
- **Published Date:** 2022-12-12
- **URL:** http://arxiv.org/abs/2212.07425v3
- **LangChain:**

   - **API Reference:** [langchain_experimental.fallacy_removal](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.fallacy_removal)

**Abstract:** The spread of misinformation, propaganda, and flawed argumentation has been
amplified in the Internet era. Given the volume of data and the subtlety of
identifying violations of argumentation norms, supporting information analytics
tasks, like content moderation, with trustworthy methods that can identify
logical fallacies is essential. In this paper, we formalize prior theoretical
work on logical fallacies into a comprehensive three-stage evaluation framework
of detection, coarse-grained, and fine-grained classification. We adapt
existing evaluation datasets for each stage of the evaluation. We employ three
families of robust and explainable methods based on prototype reasoning,
instance-based reasoning, and knowledge injection. The methods combine language
models with background knowledge and explainable mechanisms. Moreover, we
address data sparsity with strategies for data augmentation and curriculum
learning. Our three-stage framework natively consolidates prior datasets and
methods from existing tasks, like propaganda detection, serving as an
overarching evaluation testbed. We extensively evaluate these methods on our
datasets, focusing on their robustness and explainability. Our results provide
insight into the strengths and weaknesses of the methods on different
components and fallacy classes, indicating that fallacy identification is a
challenging task that may require specialized forms of reasoning to capture
various classes. We share our open-source code and data on GitHub to support
further work on logical fallacy identification.
                
## Complementary Explanations for Effective In-Context Learning

- **arXiv id:** 2211.13892v2
- **Title:** Complementary Explanations for Effective In-Context Learning
- **Authors:** Xi Ye, Srinivasan Iyer, Asli Celikyilmaz,  et al.
- **Published Date:** 2022-11-25
- **URL:** http://arxiv.org/abs/2211.13892v2
- **LangChain:**

   - **API Reference:** [langchain_core.example_selectors...MaxMarginalRelevanceExampleSelector](https://api.python.langchain.com/en/latest/example_selectors/langchain_core.example_selectors.semantic_similarity.MaxMarginalRelevanceExampleSelector.html#langchain_core.example_selectors.semantic_similarity.MaxMarginalRelevanceExampleSelector)

**Abstract:** Large language models (LLMs) have exhibited remarkable capabilities in
learning from explanations in prompts, but there has been limited understanding
of exactly how these explanations function or why they are effective. This work
aims to better understand the mechanisms by which explanations are used for
in-context learning. We first study the impact of two different factors on the
performance of prompts with explanations: the computation trace (the way the
solution is decomposed) and the natural language used to express the prompt. By
perturbing explanations on three controlled tasks, we show that both factors
contribute to the effectiveness of explanations. We further study how to form
maximally effective sets of explanations for solving a given test query. We
find that LLMs can benefit from the complementarity of the explanation set:
diverse reasoning skills shown by different exemplars can lead to better
performance. Therefore, we propose a maximal marginal relevance-based exemplar
selection approach for constructing exemplar sets that are both relevant as
well as complementary, which successfully improves the in-context learning
performance across three real-world tasks on multiple LLMs.
                
## PAL: Program-aided Language Models

- **arXiv id:** 2211.10435v2
- **Title:** PAL: Program-aided Language Models
- **Authors:** Luyu Gao, Aman Madaan, Shuyan Zhou,  et al.
- **Published Date:** 2022-11-18
- **URL:** http://arxiv.org/abs/2211.10435v2
- **LangChain:**

   - **API Reference:** [langchain_experimental.pal_chain...PALChain](https://api.python.langchain.com/en/latest/pal_chain/langchain_experimental.pal_chain.base.PALChain.html#langchain_experimental.pal_chain.base.PALChain), [langchain_experimental.pal_chain](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.pal_chain)

**Abstract:** Large language models (LLMs) have recently demonstrated an impressive ability
to perform arithmetic and symbolic reasoning tasks, when provided with a few
examples at test time ("few-shot prompting"). Much of this success can be
attributed to prompting methods such as "chain-of-thought'', which employ LLMs
for both understanding the problem description by decomposing it into steps, as
well as solving each step of the problem. While LLMs seem to be adept at this
sort of step-by-step decomposition, LLMs often make logical and arithmetic
mistakes in the solution part, even when the problem is decomposed correctly.
In this paper, we present Program-Aided Language models (PAL): a novel approach
that uses the LLM to read natural language problems and generate programs as
the intermediate reasoning steps, but offloads the solution step to a runtime
such as a Python interpreter. With PAL, decomposing the natural language
problem into runnable steps remains the only learning task for the LLM, while
solving is delegated to the interpreter. We demonstrate this synergy between a
neural LLM and a symbolic interpreter across 13 mathematical, symbolic, and
algorithmic reasoning tasks from BIG-Bench Hard and other benchmarks. In all
these natural language reasoning tasks, generating code using an LLM and
reasoning using a Python interpreter leads to more accurate results than much
larger models. For example, PAL using Codex achieves state-of-the-art few-shot
accuracy on the GSM8K benchmark of math word problems, surpassing PaLM-540B
which uses chain-of-thought by absolute 15% top-1. Our code and data are
publicly available at http://reasonwithpal.com/ .
                
## Deep Lake: a Lakehouse for Deep Learning

- **arXiv id:** 2209.10785v2
- **Title:** Deep Lake: a Lakehouse for Deep Learning
- **Authors:** Sasun Hambardzumyan, Abhinav Tuli, Levon Ghukasyan,  et al.
- **Published Date:** 2022-09-22
- **URL:** http://arxiv.org/abs/2209.10785v2
- **LangChain:**

   - **Documentation:** [docs/integrations/providers/activeloop_deeplake](https://python.langchain.com/docs/integrations/providers/activeloop_deeplake)

**Abstract:** Traditional data lakes provide critical data infrastructure for analytical
workloads by enabling time travel, running SQL queries, ingesting data with
ACID transactions, and visualizing petabyte-scale datasets on cloud storage.
They allow organizations to break down data silos, unlock data-driven
decision-making, improve operational efficiency, and reduce costs. However, as
deep learning usage increases, traditional data lakes are not well-designed for
applications such as natural language processing (NLP), audio processing,
computer vision, and applications involving non-tabular datasets. This paper
presents Deep Lake, an open-source lakehouse for deep learning applications
developed at Activeloop. Deep Lake maintains the benefits of a vanilla data
lake with one key difference: it stores complex data, such as images, videos,
annotations, as well as tabular data, in the form of tensors and rapidly
streams the data over the network to (a) Tensor Query Language, (b) in-browser
visualization engine, or (c) deep learning frameworks without sacrificing GPU
utilization. Datasets stored in Deep Lake can be accessed from PyTorch,
TensorFlow, JAX, and integrate with numerous MLOps tools.
                
## Bitext Mining Using Distilled Sentence Representations for Low-Resource Languages

- **arXiv id:** 2205.12654v1
- **Title:** Bitext Mining Using Distilled Sentence Representations for Low-Resource Languages
- **Authors:** Kevin Heffernan, Onur Çelebi, Holger Schwenk
- **Published Date:** 2022-05-25
- **URL:** http://arxiv.org/abs/2205.12654v1
- **LangChain:**

   - **API Reference:** [langchain_community.embeddings...LaserEmbeddings](https://api.python.langchain.com/en/latest/embeddings/langchain_community.embeddings.laser.LaserEmbeddings.html#langchain_community.embeddings.laser.LaserEmbeddings)

**Abstract:** Scaling multilingual representation learning beyond the hundred most frequent
languages is challenging, in particular to cover the long tail of low-resource
languages. A promising approach has been to train one-for-all multilingual
models capable of cross-lingual transfer, but these models often suffer from
insufficient capacity and interference between unrelated languages. Instead, we
move away from this approach and focus on training multiple language (family)
specific representations, but most prominently enable all languages to still be
encoded in the same representational space. To achieve this, we focus on
teacher-student training, allowing all encoders to be mutually compatible for
bitext mining, and enabling fast learning of new languages. We introduce a new
teacher-student training scheme which combines supervised and self-supervised
training, allowing encoders to take advantage of monolingual training data,
which is valuable in the low-resource setting.
  Our approach significantly outperforms the original LASER encoder. We study
very low-resource languages and handle 50 African languages, many of which are
not covered by any other model. For these languages, we train sentence
encoders, mine bitexts, and validate the bitexts by training NMT systems.
                
## Evaluating the Text-to-SQL Capabilities of Large Language Models

- **arXiv id:** 2204.00498v1
- **Title:** Evaluating the Text-to-SQL Capabilities of Large Language Models
- **Authors:** Nitarshan Rajkumar, Raymond Li, Dzmitry Bahdanau
- **Published Date:** 2022-03-15
- **URL:** http://arxiv.org/abs/2204.00498v1
- **LangChain:**

   - **API Reference:** [langchain_community.utilities...SQLDatabase](https://api.python.langchain.com/en/latest/utilities/langchain_community.utilities.sql_database.SQLDatabase.html#langchain_community.utilities.sql_database.SQLDatabase), [langchain_community.utilities...SparkSQL](https://api.python.langchain.com/en/latest/utilities/langchain_community.utilities.spark_sql.SparkSQL.html#langchain_community.utilities.spark_sql.SparkSQL)

**Abstract:** We perform an empirical evaluation of Text-to-SQL capabilities of the Codex
language model. We find that, without any finetuning, Codex is a strong
baseline on the Spider benchmark; we also analyze the failure modes of Codex in
this setting. Furthermore, we demonstrate on the GeoQuery and Scholar
benchmarks that a small number of in-domain examples provided in the prompt
enables Codex to perform better than state-of-the-art models finetuned on such
few-shot examples.
                
## Locally Typical Sampling

- **arXiv id:** 2202.00666v5
- **Title:** Locally Typical Sampling
- **Authors:** Clara Meister, Tiago Pimentel, Gian Wiher,  et al.
- **Published Date:** 2022-02-01
- **URL:** http://arxiv.org/abs/2202.00666v5
- **LangChain:**

   - **API Reference:** [langchain_community.llms...HuggingFaceTextGenInference](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference.html#langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference), [langchain_community.llms...HuggingFaceEndpoint](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint.html#langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint)

**Abstract:** Today's probabilistic language generators fall short when it comes to
producing coherent and fluent text despite the fact that the underlying models
perform well under standard metrics, e.g., perplexity. This discrepancy has
puzzled the language generation community for the last few years. In this work,
we posit that the abstraction of natural language generation as a discrete
stochastic process--which allows for an information-theoretic analysis--can
provide new insights into the behavior of probabilistic language generators,
e.g., why high-probability texts can be dull or repetitive. Humans use language
as a means of communicating information, aiming to do so in a simultaneously
efficient and error-minimizing manner; in fact, psycholinguistics research
suggests humans choose each word in a string with this subconscious goal in
mind. We formally define the set of strings that meet this criterion: those for
which each word has an information content close to the expected information
content, i.e., the conditional entropy of our model. We then propose a simple
and efficient procedure for enforcing this criterion when generating from
probabilistic models, which we call locally typical sampling. Automatic and
human evaluations show that, in comparison to nucleus and top-k sampling,
locally typical sampling offers competitive performance (in both abstractive
summarization and story generation) in terms of quality while consistently
reducing degenerate repetitions.
                
## Learning Transferable Visual Models From Natural Language Supervision

- **arXiv id:** 2103.00020v1
- **Title:** Learning Transferable Visual Models From Natural Language Supervision
- **Authors:** Alec Radford, Jong Wook Kim, Chris Hallacy,  et al.
- **Published Date:** 2021-02-26
- **URL:** http://arxiv.org/abs/2103.00020v1
- **LangChain:**

   - **API Reference:** [langchain_experimental.open_clip](https://api.python.langchain.com/en/latest/experimental_api_reference.html#module-langchain_experimental.open_clip)

**Abstract:** State-of-the-art computer vision systems are trained to predict a fixed set
of predetermined object categories. This restricted form of supervision limits
their generality and usability since additional labeled data is needed to
specify any other visual concept. Learning directly from raw text about images
is a promising alternative which leverages a much broader source of
supervision. We demonstrate that the simple pre-training task of predicting
which caption goes with which image is an efficient and scalable way to learn
SOTA image representations from scratch on a dataset of 400 million (image,
text) pairs collected from the internet. After pre-training, natural language
is used to reference learned visual concepts (or describe new ones) enabling
zero-shot transfer of the model to downstream tasks. We study the performance
of this approach by benchmarking on over 30 different existing computer vision
datasets, spanning tasks such as OCR, action recognition in videos,
geo-localization, and many types of fine-grained object classification. The
model transfers non-trivially to most tasks and is often competitive with a
fully supervised baseline without the need for any dataset specific training.
For instance, we match the accuracy of the original ResNet-50 on ImageNet
zero-shot without needing to use any of the 1.28 million training examples it
was trained on. We release our code and pre-trained model weights at
https://github.com/OpenAI/CLIP.
                
## CTRL: A Conditional Transformer Language Model for Controllable Generation

- **arXiv id:** 1909.05858v2
- **Title:** CTRL: A Conditional Transformer Language Model for Controllable Generation
- **Authors:** Nitish Shirish Keskar, Bryan McCann, Lav R. Varshney,  et al.
- **Published Date:** 2019-09-11
- **URL:** http://arxiv.org/abs/1909.05858v2
- **LangChain:**

   - **API Reference:** [langchain_community.llms...HuggingFaceTextGenInference](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference.html#langchain_community.llms.huggingface_text_gen_inference.HuggingFaceTextGenInference), [langchain_community.llms...HuggingFaceEndpoint](https://api.python.langchain.com/en/latest/llms/langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint.html#langchain_community.llms.huggingface_endpoint.HuggingFaceEndpoint)

**Abstract:** Large-scale language models show promising text generation capabilities, but
users cannot easily control particular aspects of the generated text. We
release CTRL, a 1.63 billion-parameter conditional transformer language model,
trained to condition on control codes that govern style, content, and
task-specific behavior. Control codes were derived from structure that
naturally co-occurs with raw text, preserving the advantages of unsupervised
learning while providing more explicit control over text generation. These
codes also allow CTRL to predict which parts of the training data are most
likely given a sequence. This provides a potential method for analyzing large
amounts of data via model-based source attribution. We have released multiple
full-sized, pretrained versions of CTRL at https://github.com/salesforce/ctrl.
                
## Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks

- **arXiv id:** 1908.10084v1
- **Title:** Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
- **Authors:** Nils Reimers, Iryna Gurevych
- **Published Date:** 2019-08-27
- **URL:** http://arxiv.org/abs/1908.10084v1
- **LangChain:**

   - **Documentation:** [docs/integrations/text_embedding/sentence_transformers](https://python.langchain.com/docs/integrations/text_embedding/sentence_transformers)

**Abstract:** BERT (Devlin et al., 2018) and RoBERTa (Liu et al., 2019) has set a new
state-of-the-art performance on sentence-pair regression tasks like semantic
textual similarity (STS). However, it requires that both sentences are fed into
the network, which causes a massive computational overhead: Finding the most
similar pair in a collection of 10,000 sentences requires about 50 million
inference computations (~65 hours) with BERT. The construction of BERT makes it
unsuitable for semantic similarity search as well as for unsupervised tasks
like clustering.
  In this publication, we present Sentence-BERT (SBERT), a modification of the
pretrained BERT network that use siamese and triplet network structures to
derive semantically meaningful sentence embeddings that can be compared using
cosine-similarity. This reduces the effort for finding the most similar pair
from 65 hours with BERT / RoBERTa to about 5 seconds with SBERT, while
maintaining the accuracy from BERT.
  We evaluate SBERT and SRoBERTa on common STS tasks and transfer learning
tasks, where it outperforms other state-of-the-art sentence embeddings methods.