9). TensorLLM Proposes a framework that performs MHA compression through a multi-head tensorisation process and the Tucker decomposition. Achieves a compression rate of up to ∼ 250x in the MHA weights…
@dair_ai
-
TokenVerse Introduces Novel Image Generation from Learned Concepts
By
–
10). TokenVerse Proposes a new technique to generate new images from learned concepts in a desired configuration.
-
Improving RAG Through Multi-Agent Reinforcement Learning
By
–
8). Improving RAG through Multi-Agent RL It models RAG components like query rewriting, document selection, and answer generation as RL agents working together toward generating accurate answers.
-
Docling: Open-Source Document Parsing Toolkit
By
–
7). Docling Docling is an open-source toolkit that can parse several types of popular document formats into a unified, richly structured representation. https://
arxiv.org/abs/2501.17887 -
DeepSeek-R1 Usage Recommendations and Prompting Guide
By
–
6). Usage Recommendation for DeepSeek-R1 This work provides a set of recommendations for how to prompt the DeepSeek-R1 model.
-
Diverse Preference Optimization: Novel Training Method for Language Models
By
–
5). Diverse Preference Optimization A novel training method that aims to address the lack of diversity in language model outputs while maintaining response quality.
-
O1-like LLMs underthinking patterns and limitations
By
–
4). On the Underthinking of o1-like LLMs Looks more closely at the "thinking" patterns of o1-like LLMs. We have seen a few recent papers pointing out the issues with overthinking.
-
Janus-Pro: Enhanced Multimodal Understanding and Generation Model
By
–
3). Janus-Pro An enhanced version of the previous Janus model for multimodal understanding and generation.
-
Qwen Releases 1M Token Context Open-Source LLMs
By
–
2). Qwen2.5-1M Qwen releases two open-source LLMs, Qwen2.5-7B-Instruct-1M and Qwen2.5-14B-Instruct-1M, that can handle context lengths of up to 1 million tokens.
-
ChemAgent Framework Enhances LLM Chemical Reasoning Performance
By
–
10). ChemAgent – presents a new framework designed to improve the performance of LLMs on chemical reasoning through a dynamic, self-updating library…