Large Language Models

Personalized Graph-based Retrieval-Augmented Generation (PGraphRAG) significantly improves personalized text generation by leveraging user-centric knowledge graphs, especially in cold-start scenarios …

METAGENE-1, a 7-billion parameter language model, achieves state-of-the-art results in pathogen detection and genomic embedding by leveraging a massive wastewater metagenomic dataset.

AUTO-RT automates LLM vulnerability discovery by using reinforcement learning to optimize complex attack strategies, achieving faster detection and higher success rates than existing methods.

Boosting code generation accuracy with more unit tests! This research shows that increasing the number of unit tests used to evaluate code generated by LLMs significantly improves accuracy, especially…

CODEELO benchmark uses CodeForces to fairly evaluate LLMs’ coding abilities, providing human-comparable Elo ratings and addressing limitations of existing benchmarks.

BoxingGym: A new benchmark rigorously evaluates AI agents’ ability to design experiments and discover scientific models, revealing current LLMs’ limitations and highlighting fertile research avenues.

LUSIFER: a novel zero-shot approach empowers English-centric LLM embedding models for multilingual tasks without explicit multilingual training data, significantly enhancing performance, especially fo…

Polarizing SSMs’ state transition matrices enhances long-range dependency modeling by mitigating recency bias and over-smoothing.

New benchmarks, HumanEval Pro and MBPP Pro, reveal LLMs struggle with self-invoking code generation, highlighting a critical gap in current code reasoning capabilities.

Researchers introduce Learned Embedding Propagation (LEP), a novel technique that efficiently adapts large language models (LLMs) to new languages using minimal training data, thus overcoming limitati…

Dynasor optimizes LLM reasoning by dynamically allocating compute based on a novel ‘certaindex’ metric, reducing compute by up to 50% and increasing query rates by 3.3x.

Xmodel-2: A 1.2B parameter LLM achieving state-of-the-art reasoning performance through efficient architecture and training, now publicly available!

Boost fine-tuned LLMs’ performance without sacrificing safety by merging pre- and post-tuning model weights!

TALE: A novel framework dynamically adjusts token budgets in LLM reasoning prompts, slashing costs by ~70% with minimal accuracy loss.

Molar: A novel multimodal LLM framework boosts sequential recommendation accuracy by cleverly aligning collaborative filtering with rich item representations from text and non-text data.

YuLan-Mini: An open, data-efficient 2.42B parameter LLM achieving top-tier performance with innovative training techniques.

LLM evaluation on multiple-choice questions is flawed; considering all options simultaneously, not individually, reveals much higher accuracy and challenges existing benchmark rankings.

FoPE enhances attention’s periodic extension for better length generalization in language models by addressing spectral damage in RoPE using Fourier Series and zeroing out destructive frequencies.

Frozen LLMs get a performance boost by augmenting their key-value cache with latent embeddings generated by a differentiable offline coprocessor.

B-STAR dynamically balances exploration and exploitation in self-taught reasoners, achieving superior performance in mathematical, coding, and commonsense reasoning tasks.