arxiv_id stringlengths 10 10 | title stringlengths 8 162 | summary stringlengths 228 1.92k | pdf stringlengths 14 14 | authors listlengths 1 451 | categories listlengths 1 6 | primary_category stringclasses 40
values | published timestamp[s]date 2015-02-08 13:57:59 2024-01-08 18:47:34 | updated timestamp[s]date 2015-03-02 20:44:12 2026-06-12 08:15:52 | references null | content stringlengths 8.83k 951k ⌀ |
|---|---|---|---|---|---|---|---|---|---|---|
2401.04088 | Mixtral of Experts | We introduce Mixtral 8x7B, a Sparse Mixture of Experts (SMoE) language model. Mixtral has the same architecture as Mistral 7B, with the difference that each layer is composed of 8 feedforward blocks (i.e. experts). For every token, at each layer, a router network selects two experts to process the current state and com... | 2401.04088.pdf | [
"Albert Q. Jiang",
"Alexandre Sablayrolles",
"Antoine Roux",
"Arthur Mensch",
"Blanche Savary",
"Chris Bamford",
"Devendra Singh Chaplot",
"Diego de las Casas",
"Emma Bou Hanna",
"Florian Bressand",
"Gianna Lengyel",
"Guillaume Bour",
"Guillaume Lample",
"Lélio Renard Lavaud",
"Lucile Sa... | [
"cs.LG",
"cs.CL"
] | cs.LG | 2024-01-08T18:47:34 | 2024-01-08T18:47:34 | null | ## Mixtral of Experts
Albert Q. Jiang, Alexandre Sablayrolles, Antoine Roux, Arthur Mensch, Blanche Savary, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Emma Bou Hanna, Florian Bressand, Gianna Lengyel, Guillaume Bour, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Sto... |
2312.17238 | Fast Inference of Mixture-of-Experts Language Models with Offloading | With the widespread adoption of Large Language Models (LLMs), many deep learning practitioners are looking for strategies of running these models more efficiently. One such strategy is to use sparse Mixture-of-Experts (MoE) - a type of model architectures where only a fraction of model layers are active for any given i... | 2312.17238.pdf | [
"Artyom Eliseev",
"Denis Mazur"
] | [
"cs.LG",
"cs.AI",
"cs.DC"
] | cs.LG | 2023-12-28T18:58:13 | 2023-12-28T18:58:13 | null | ## Fast Inference of Mixture-of-Experts Language Models with Offloading
## Artyom Eliseev
Moscow Institute of Physics and Technology Yandex School of Data Analysis lavawolfiee@gmail.com
## Denis Mazur
Moscow Institute of Physics and Technology Yandex Researchcore denismazur8@gmail.com
## Abstract
With the widespr... |
2312.11111 | The Good, The Bad, and Why: Unveiling Emotions in Generative AI | "Emotion significantly impacts our daily behaviors and interactions. While recent generative AI mode(...TRUNCATED) | 2312.11111.pdf | ["Cheng Li","Jindong Wang","Yixuan Zhang","Kaijie Zhu","Xinyi Wang","Wenxin Hou","Jianxun Lian","Fan(...TRUNCATED) | [
"cs.AI",
"cs.CL",
"cs.HC"
] | cs.AI | 2023-12-18T11:19:45 | 2024-06-07T09:25:31 | null | "## The Good, The Bad, and Why: Unveiling Emotions in Generative AI\n\nCheng Li 1 2 Jindong Wang 1 Y(...TRUNCATED) |
2312.00752 | Mamba: Linear-Time Sequence Modeling with Selective State Spaces | "Foundation models, now powering most of the exciting applications in deep learning, are almost univ(...TRUNCATED) | 2312.00752.pdf | [
"Albert Gu",
"Tri Dao"
] | [
"cs.LG",
"cs.AI"
] | cs.LG | 2023-12-01T18:01:34 | 2024-05-31T17:55:27 | null | "## Mamba: Linear-Time Sequence Modeling with Selective State Spaces\n\nAlbert Gu ∗ 1 and Tri Dao (...TRUNCATED) |
2311.15296 | "UHGEval: Benchmarking the Hallucination of Chinese Large Language Models via Unconstrained Generati(...TRUNCATED) | "Large language models (LLMs) have emerged as pivotal contributors in contemporary natural language (...TRUNCATED) | 2311.15296.pdf | ["Xun Liang","Shichao Song","Simin Niu","Zhiyu Li","Feiyu Xiong","Bo Tang","Yezhaohui Wang","Dawei H(...TRUNCATED) | [
"cs.CL"
] | cs.CL | 2023-11-26T13:42:56 | 2024-05-24T03:29:14 | null | "## U H G E v l a UHG Eval: Benchmarking the Hallucination of Chinese Large Language Models via Unco(...TRUNCATED) |
2311.04254 | Everything of Thoughts: Defying the Law of Penrose Triangle for Thought Generation | "Recent advancements in Large Language Models (LLMs) have revolutionized decision-making by breaking(...TRUNCATED) | 2311.04254.pdf | ["Ruomeng Ding","Chaoyun Zhang","Lu Wang","Yong Xu","Minghua Ma","Wei Zhang","Si Qin","Saravan Rajmo(...TRUNCATED) | [
"cs.AI",
"cs.LG"
] | cs.AI | 2023-11-07T12:30:36 | 2024-02-23T15:09:58 | null | "## EVERYTHING OF THOUGHTS : DEFYING THE LAW OF PENROSE TRIANGLE FOR THOUGHT GENERATION\n\nRuomeng D(...TRUNCATED) |
2311.04072 | Beyond Imitation: Leveraging Fine-grained Quality Signals for Alignment | "Alignment with human preference is a desired property of large language models (LLMs). Currently, t(...TRUNCATED) | 2311.04072.pdf | [
"Geyang Guo",
"Ranchi Zhao",
"Tianyi Tang",
"Wayne Xin Zhao",
"Ji-Rong Wen"
] | [
"cs.CL"
] | cs.CL | 2023-11-07T15:36:40 | 2024-04-15T15:25:53 | null | "## BEYOND IMITATION: LEVERAGING FINE-GRAINED QUALITY SIGNALS FOR ALIGNMENT\n\nGeyang Guo 1 ∗ , Ra(...TRUNCATED) |
2311.01964 | Don't Make Your LLM an Evaluation Benchmark Cheater | "Large language models~(LLMs) have greatly advanced the frontiers of artificial intelligence, attain(...TRUNCATED) | 2311.01964.pdf | ["Kun Zhou","Yutao Zhu","Zhipeng Chen","Wentong Chen","Wayne Xin Zhao","Xu Chen","Yankai Lin","Ji-Ro(...TRUNCATED) | [
"cs.CL",
"cs.AI"
] | cs.CL | 2023-11-03T14:59:54 | 2023-11-03T14:59:54 | null | "## Don't Make Your LLM an Evaluation Benchmark Cheater\n\nKun Zhou 1 , Yutao Zhu 2 , Zhipeng Chen 2(...TRUNCATED) |
2311.04915 | "Chain of Empathy: Enhancing Empathetic Response of Large Language Models Based on Psychotherapy Mod(...TRUNCATED) | "We present a novel method, the Chain of Empathy (CoE) prompting, that utilizes insights from psycho(...TRUNCATED) | 2311.04915.pdf | [
"Yoon Kyung Lee",
"Inju Lee",
"Minjung Shin",
"Seoyeon Bae",
"Sowon Hahn"
] | [
"cs.CL",
"cs.AI",
"cs.HC"
] | cs.CL | 2023-11-02T02:21:39 | 2024-09-14T04:49:08 | null | "## Chain of Empathy: Enhancing Empathetic Response of Large Language Models Based on Psychotherapy (...TRUNCATED) |
2311.01555 | Instruction Distillation Makes Large Language Models Efficient Zero-shot Rankers | "Recent studies have demonstrated the great potential of Large Language Models (LLMs) serving as zer(...TRUNCATED) | 2311.01555.pdf | ["Weiwei Sun","Zheng Chen","Xinyu Ma","Lingyong Yan","Shuaiqiang Wang","Pengjie Ren","Zhumin Chen","(...TRUNCATED) | [
"cs.IR",
"cs.CL"
] | cs.IR | 2023-11-02T19:16:21 | 2023-11-02T19:16:21 | null | "## Instruction Distillation Makes Large Language Models Efficient Zero-shot Rankers\n\nWeiwei Sun 1(...TRUNCATED) |
End of preview. Expand in Data Studio
AI ArXiv Papers
A reprocessed version of the paper collection from jamescalam/ai-arxiv2, independently re-downloaded from the arXiv API and converted from PDF to Markdown using Docling.
Relationship to the Original Dataset
This dataset reconstructs the paper collection represented by jamescalam/ai-arxiv2, but does not reuse its PDF-to-text processing.
Papers were retrieved independently from arXiv and parsed using Docling. As a result, the Markdown content may differ from the original dataset's text due to:
- PDF versions available at retrieval time
- Differences in PDF parsing and document-structure extraction
- Markdown conversion and formatting
- Handling of equations, tables, figures, captions, and references
- Document metadata differences
- Changes to papers between arXiv versions
Fields
| Field | Description |
|---|---|
arxiv_id |
arXiv identifier |
title |
Paper title |
summary |
arXiv abstract/summary |
pdf |
PDF filename/reference |
authors |
List of authors |
categories |
arXiv categories |
primary_category |
Primary arXiv category |
published |
Original publication timestamp |
updated |
Last update timestamp |
references |
Reference information, if present |
content |
Full paper converted from PDF to Markdown using Docling |
Data Processing
- Paper identifiers were obtained from the original
jamescalam/ai-arxiv2dataset. - The corresponding papers were downloaded independently via the arXiv API.
- The downloaded PDFs were converted to Markdown using Docling.
- The resulting Markdown content was combined with the corresponding metadata into the final dataset records.
Intended Use
This dataset may support research and experimentation in:
- Retrieval-Augmented Generation (RAG)
- Scientific document retrieval and semantic search
- Document understanding and parsing evaluation
- Long-context language model experiments
- Scientific NLP and information extraction
- Citation and reference extraction
- Embeddings and vector search
- Training or evaluating systems for scientific literature understanding
Data Quality
The Markdown content is generated automatically from PDFs and should not be assumed to perfectly represent the original documents.
Potential parsing issues include:
- Mathematical notation and equations
- Tables and multi-column layouts
- Figures and captions
- Footnotes, headers, and footers
- Page boundaries
- References
- Reading order
For high-precision research, consult the original paper/PDF when exact wording, equations, tables, or formatting matter.
Limitations
- PDF parsing — Automated conversion may introduce errors even when the source PDF is well formatted.
- Version differences — The retrieved PDF may differ from the arXiv version used elsewhere.
- Metadata — Metadata reflects the information available at the time of collection.
- Coverage — This dataset represents the specific collection reconstructed from
jamescalam/ai-arxiv2; it is not a complete snapshot of AI or computer-science papers on arXiv. - No manual review — The Markdown documents have not necessarily been manually inspected or corrected.
Licensing and Copyright
The papers originate from arXiv, and the applicable copyright and licensing terms may vary between individual papers. No single license necessarily applies to the underlying paper content across the entire dataset.
The PDF-to-Markdown conversion performed using Docling is a transformation of the source documents and does not itself change the copyright or licensing status of the underlying papers.
Users are responsible for determining the applicable license and copyright conditions for individual papers and for ensuring that their use of the material complies with those conditions.
Source
- Original dataset:
jamescalam/ai-arxiv2 - Paper source: arXiv
- Document conversion: Docling
Citation
If you use this dataset, please cite this dataset and the original papers used in your work.
For individual papers, please cite the corresponding arXiv entry and its authors.
Acknowledgements
- arXiv — for access to the underlying research papers and metadata
- Docling — for PDF parsing and Markdown conversion
- jamescalam/ai-arxiv2 — as the reference collection used to reconstruct this paper set
- Downloads last month
- 68