4). DataGemma – includes a series of fine-tuned Gemma 2 models to help LLMs access and incorporate numerical and statistical data; proposes a new approach called Retrieval Interleaved Generation (RIG) which can reliably incorporate public statistical data from Data Commons into
@dair_ai
-
Chai-1: Multimodal Foundation Model for Molecular Structure Prediction
By
–
2). Chai-1 – a new multi-modal foundation model for molecular structure prediction that can predict proteins, small molecules, DNA, RNA, and more; it achieves state-of-the-art results on a variety of tasks in drug discovery…
-
LLMs Learn to Reason with Reinforcement Learning Training
By
–
1). Learning to Reason with LLMs – a new family of LLMs trained with reinforcement learning to reason before it responds to complex tasks; it produces a long internal chain of thought and exceeds in science, code, and math-related tasks…
-
Top ML Papers of the Week: LLMs, Agents, and Research
By
–
The Top ML Papers of the Week (Sep 9 – Sep 15): – Chai-1
– DataGemma
– LLaMa-Omni
– Agent Workflow Memory
– Learning to Reason with LLMs
– Can LLMs Generation Novel Research Ideas Read on for more: -
Beyond Human Preference: Reframing AI Alignment Challenges
By
–
9). Beyond Preference in AI Alignment – challenges the dominant practice of AI alignment known as human preference tuning; explains in what ways human preference tuning fails to capture the thick semantic content of human values; argues that AI alignment needs reframing, instead
-
RAG Noise Benchmark Measures LLM Performance Impact
By
–
8). Role of RAG Noise in LLMs – proposes a benchmark (NoiserBench) to measure how different kinds of noisy information affect RAG's performance; reports that from different kinds of beneficial noise studied (e.g., semantic, datatype, and illegal sentence), illegal sentence noise
-
LLM-Based Agents for Software Engineering: Comprehensive Survey
By
–
10). LLM-Based Agents for Software Engineering – a survey paper on LLM-based agents for software engineering, covering perspectives ranging from requirement engineering to test generation to software maintenance.
-
LongCite Improves Long-Context Question Answering With Citations
By
–
6). LongCite – synthesizes a large-scale SFT dataset with off-the-shelf LLMs to improve long-context question answering with citations; it trains 8B and 9B parameter models that enhance citation generation capabilities from lengthy contexts while improving response correctness;
-
MemLong extends LLM context length using external retrieval
By
–
7). MemLong – utilizes an external retriever for retrieving historical information which enhances the capabilities of long-context LLMs; it consistently outperforms other SoTA LLMs on long-context benchmarks and can extend the context length on a single 3090 GPU from 4k up to
-
OLMoE: Open Sparse Mixture-of-Experts Language Model
By
–
5). OLMoE – introduces a fully-open LLM that leverages sparse Mixture-of-Experts. OLMoE is a 7B parameter model and uses 1B active parameters per input token; there is also an instruction-tuned version that claims to outperform Llama-2-13B-Chat and DeepSeekMoE 16B.