| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | AA | AB | AC | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Type | Date | Title | Image | Link | Content | |||||||||||||||||||||||
2 | Paper | 1/09/2023 | RLAIF: Scaling Reinforcement Learning from Human Feedback with AI Feedback | https://arxiv.org/abs/2309.00267 | RLAIF's AI-driven feedback matches human feedback in aligning LLMs, with both outperforming the baseline in 70% of tests, offering a scalable alternative to RLHF. | ||||||||||||||||||||||||
3 | Model | 04/09/2023 | TinyLlama | https://github.com/jzhang38/TinyLlama/raw/main/.github/TinyLlama_logo.png | https://github.com/jzhang38/TinyLlama | A 1.1B-parameter Llama 2 clone pretrained on 3T tokens in 90 days with 16 GPUs—compact power for projects needing low compute footprint. | |||||||||||||||||||||||
4 | Paper | 05/09/2023 | Cognitive Architecture for Language Agents | https://arxiv.org/abs/2309.02427 | Drawing from symbolic AI's agent design, they introduce the CoALA framework, blending LLMs with cognitive architectures for systematic development of advanced language agents. | ||||||||||||||||||||||||
5 | Model | 6/09/2023 | Falcon 180B | https://encrypted-tbn2.gstatic.com/images?q=tbn:ANd9GcR3EnofDsPdbMmBW47p610_zQzX8kNUzzUZx0DnjaPNkIlWcKw6 | https://huggingface.co/blog/falcon-180b | Abu Dhabi's Tech Innovation Institute introduces Falcon 180B, a open-use top-ranking language model rivaling GPT-4 and Google's PaLM 2 Large. | |||||||||||||||||||||||
6 | Paper | 06/09/2023 | GPT Can Solve Mathematical Problems Without a Calculator | https://arxiv.org/abs/2309.03241 | A 2 billion-parameter model outperforms GPT-4 in multi-digit arithmetic if trained with sufficient data and matches GPT-4 in Chinese math problems when fine-tuned. | ||||||||||||||||||||||||
7 | Paper | 7/09/2023 | Large Language Models as Optimizers | https://arxiv.org/abs/2309.03409 | Optimization by PROmpting (OPRO) method uses LLMs to optimize tasks described in natural language, surpassing human prompts by up to 50%. | ||||||||||||||||||||||||
8 | Model | 7/09/2023 | FLM-101B | https://huggingface.co/CofeAI/FLM-101B | 101B-parameter LLM trained under $100K budget matches GPT-3 performance with innovative IQ-evaluation. | ||||||||||||||||||||||||
9 | News | 8/09/2023 | TensorRT-LLM | NVIDIA TensorRT-LLM Supercharges Large Language Model Inference on NVIDIA H100 GPUs | NVIDIA Technical Blog | NVIDIA unveils TensorRT-LLM for optimizing large language models, supports popular LLMs and it will be integrated into the NVIDIA NeMo framework. | ||||||||||||||||||||||||
10 | Paper | 8/09/2023 | From Sparse to Dense: GPT-4 Summarization with Chain of Density Prompting | https://arxiv.org/abs/2309.04269 | The "Chain of Density" (CoD) prompt enables GPT-4 to produce denser summaries preferred by humans, balancing informativeness and readability. | ||||||||||||||||||||||||
11 | Model | 9/09/2023 | Baichuan 2 | https://huggingface.co/baichuan-inc | Baichuan 2, a large-scale multilingual language model with up to 13 billion parameters, outperforms similar open-source models on key benchmarks and excels in domains like medicine and law. | ||||||||||||||||||||||||
12 | Paper | 11/09/2023 | NExT-GPT: Any-to-Any Multimodal LLM | https://arxiv.org/abs/2309.05519 | NExT-GPT enables any-to-any content generation across text, images, videos, and audio with minimal parameter tuning of an only-text LLM, advancing human-like AI interaction. | ||||||||||||||||||||||||
13 | Paper | 11/09/2023 | Textbooks Are All You Need II: phi-1.5 technical report | https://arxiv.org/abs/2309.05463 | The study unveils phi-1.5, a 1.3 billion parameter model that matches larger models in performance, using "textbook quality" data for improved reasoning and is open-sourced for further exploration. | ||||||||||||||||||||||||
14 | Paper | 12/09/2023 | A Survey of Hallucination in Large Foundation Models | https://arxiv.org/abs/2309.05922 | This survey dives deep into the concept of 'hallucination' in LLMs, classifying its types, establishing evaluation standards, reviewing mitigation strategies, and charting the path for future research. | ||||||||||||||||||||||||
15 | News | 13/09/2023 | Open Interpreter | https://github.com/KillianLucas/open-interpreter/ | A permisively licensed alternative to OpenAI's CodeInterpreter that allows diferent LLMs to run locally and use natural language to run code. | ||||||||||||||||||||||||
16 | Paper | 15/09/2023 | Connecting Large Language Models with Evolutionary Algorithms Yields Powerful Prompt Optimizers | https://arxiv.org/abs/2309.08532 | The paper introduces "EvoPrompt", a framework using evolutionary algorithms to optimize prompts for LLMs, yielding significant improvements over human-crafted and other automatic methods. | ||||||||||||||||||||||||
17 | Paper | 17/09/2023 | Contrastive Decoding Improves Reasoning in Large Language Models | https://arxiv.org/abs/2309.09117 | Contrastive Decoding improves over existing methods by preventing some abstract reasoning errors, as well as by avoiding simpler modes such as copying sections of the input during chain-of-thought. | ||||||||||||||||||||||||
18 | Paper | 18/09/2023 | Adapting Large Language Models via Reading Comprehension | https://arxiv.org/abs/2309.09530 | By transforming raw corpora into reading comprehension texts, domain-specific training for LLMs is improved, achieving competitive performance across multiple domains, even rivaling larger specialized models. | ||||||||||||||||||||||||
19 | Paper | 19/09/2023 | LMDX: Language Model-based Document Information Extraction and Localization | https://arxiv.org/abs/2309.10952 | LMDX is a novel method to adapt LLMs for extracting and localizing information from visually rich documents, surpassing previous benchmarks and ensuring data efficiency. | ||||||||||||||||||||||||
20 | Paper | 19/09/2023 | Language Modeling Is Compression | https://arxiv.org/abs/2309.10668 | LLMs, due to their predictive prowess, can be powerful compressors; Chinchilla 70B outperforms domain-specific compressors, highlighting the interconnectedness of prediction and compression. | ||||||||||||||||||||||||
21 | Paper | 20/09/2023 | Chain-of-Verification Reduces Hallucination in Large Language Models | https://arxiv.org/abs/2309.11495 | Chain-of-Verification makes the model fact check its response iteratively reducing the amount of plausible yet incorrect factual information that it produces. | ||||||||||||||||||||||||
22 | Paper | 20/09/2023 | A Paradigm Shift in Machine Translation: Boosting Translation Performance of Large Language Models | https://arxiv.org/abs/2309.11674 | Using a two-stage fine-tuning process on both monolingual and high-quality parallel data significantly improves translation performance of moderate-sized LLMs, outperforming even larger models. | ||||||||||||||||||||||||
23 | Model | 20/09/2023 | BTLM-3B-8K | https://arxiv.org/abs/2309.11568 | BTLM-3B-8K sets a new standard, outperforming other 3B models and even rivaling some 7B models; optimized for mobile and edge devices, it's available under Apache 2.0. | ||||||||||||||||||||||||
24 | Paper | 21/09/2023 | LongLoRA: Efficient Fine-tuning of Long-Context Large Language Models | https://arxiv.org/abs/2309.12307 | An efficient fine-tuning approach that extends the context sizes of pre-trained LLMs reducing the computational costs of training for longer context lengths. | ||||||||||||||||||||||||
25 | Paper | 23/09/2023 | Exploring Large Language Models' Cognitive Moral Development through Defining Issues Test | https://arxiv.org/abs/2309.13356 | This study builds a bridge between the fields of psychology and AI evaluating LLMs' ethical reasoning using psychology principles and the Psychometric Assessment Tool-Defining Issues Test. | ||||||||||||||||||||||||
26 | Model | 25/09/2023 | Qwen | https://github.com/QwenLM/Qwen/blob/main/assets/logo.jpg | https://github.com/QwenLM/Qwen | 7B and14B parameter models that improve the current state of the art models of similar size. Allows for context lenghts up to 8192. | |||||||||||||||||||||||
27 | Paper | 26/09/2023 | QA-LoRA: Quantization-Aware Low-Rank Adaptation of Large Language Models | https://arxiv.org/abs/2309.14717 | The paper creates QA-LoRA which introduces quantization to the fine-tuning process to get less acccuracy loss while reducing training time and memory costs. | ||||||||||||||||||||||||
28 | Model | 27/09/2023 | Mistral 7B | https://techcrunch.com/wp-content/uploads/2023/09/mistral-7b-v0.1.jpg?w=990&crop=1 | https://huggingface.co/mistralai/Mistral-7B-v0.1 | One of the best performing 7B parameter models to date, outperforming bigger models like Llama 2 13B on all benchmarks. Released under Apache 2.0 license so it can be used without restrictions. | |||||||||||||||||||||||
29 | Paper | 27/09/2023 | Effective Long-Context Scaling of Foundation Models | https://arxiv.org/abs/2309.16039 | Meta introduces Llama-2-Long with up to 32,768 tokens context window, outperforming previous models on long-context tasks through a unique pretraining approach. | ||||||||||||||||||||||||
30 | News | 28/09/2023 | Amazon Bedrock | https://upload.wikimedia.org/wikipedia/commons/thumb/9/93/Amazon_Web_Services_Logo.svg/2560px-Amazon_Web_Services_Logo.svg.png | https://aws.amazon.com/bedrock/ | Amazon Bedrock is a fully managed service that offers a choice of high-performing foundation models from leading AI companies, simplifying the development of generative AI applications. | |||||||||||||||||||||||
31 | |||||||||||||||||||||||||||||
32 | |||||||||||||||||||||||||||||
33 | |||||||||||||||||||||||||||||
34 | |||||||||||||||||||||||||||||
35 | |||||||||||||||||||||||||||||
36 | |||||||||||||||||||||||||||||
37 | |||||||||||||||||||||||||||||
38 | |||||||||||||||||||||||||||||
39 | |||||||||||||||||||||||||||||
40 | |||||||||||||||||||||||||||||
41 | |||||||||||||||||||||||||||||
42 | |||||||||||||||||||||||||||||
43 | |||||||||||||||||||||||||||||
44 | |||||||||||||||||||||||||||||
45 | |||||||||||||||||||||||||||||
46 | |||||||||||||||||||||||||||||
47 | |||||||||||||||||||||||||||||
48 | |||||||||||||||||||||||||||||
49 | |||||||||||||||||||||||||||||
50 | |||||||||||||||||||||||||||||
51 | |||||||||||||||||||||||||||||
52 | |||||||||||||||||||||||||||||
53 | |||||||||||||||||||||||||||||
54 | |||||||||||||||||||||||||||||
55 | |||||||||||||||||||||||||||||
56 | |||||||||||||||||||||||||||||
57 | |||||||||||||||||||||||||||||
58 | |||||||||||||||||||||||||||||
59 | |||||||||||||||||||||||||||||
60 | |||||||||||||||||||||||||||||
61 | |||||||||||||||||||||||||||||
62 | |||||||||||||||||||||||||||||
63 | |||||||||||||||||||||||||||||
64 | |||||||||||||||||||||||||||||
65 | |||||||||||||||||||||||||||||
66 | |||||||||||||||||||||||||||||
67 | |||||||||||||||||||||||||||||
68 | |||||||||||||||||||||||||||||
69 | |||||||||||||||||||||||||||||
70 | |||||||||||||||||||||||||||||
71 | |||||||||||||||||||||||||||||
72 | |||||||||||||||||||||||||||||
73 | |||||||||||||||||||||||||||||
74 | |||||||||||||||||||||||||||||
75 | |||||||||||||||||||||||||||||
76 | |||||||||||||||||||||||||||||
77 | |||||||||||||||||||||||||||||
78 | |||||||||||||||||||||||||||||
79 | |||||||||||||||||||||||||||||
80 | |||||||||||||||||||||||||||||
81 | |||||||||||||||||||||||||||||
82 | |||||||||||||||||||||||||||||
83 | |||||||||||||||||||||||||||||
84 | |||||||||||||||||||||||||||||
85 | |||||||||||||||||||||||||||||
86 | |||||||||||||||||||||||||||||
87 | |||||||||||||||||||||||||||||
88 | |||||||||||||||||||||||||||||
89 | |||||||||||||||||||||||||||||
90 | |||||||||||||||||||||||||||||
91 | |||||||||||||||||||||||||||||
92 | |||||||||||||||||||||||||||||
93 | |||||||||||||||||||||||||||||
94 | |||||||||||||||||||||||||||||
95 | |||||||||||||||||||||||||||||
96 | |||||||||||||||||||||||||||||
97 | |||||||||||||||||||||||||||||
98 | |||||||||||||||||||||||||||||
99 | |||||||||||||||||||||||||||||
100 |