2025-01-31s

WILDCHAT-50M: Largest public chat dataset refines LLM post-training, showing superior SFT performance with fewer samples.

Large language models (LLMs) often prematurely abandon promising reasoning paths, a phenomenon called ‘underthinking’. This paper introduces a novel metric to quantify this issue and proposes a decodi…

Streaming DiLoCo achieves two orders of magnitude bandwidth reduction in billion-scale parameter LLM training by synchronizing parameter subsets sequentially, overlapping communication with computatio…

ASTRAL, a novel automated safety testing tool, reveals DeepSeek-R1’s significantly higher unsafe response rate compared to OpenAI’s o3-mini, highlighting critical safety concerns in advanced LLMs.

GuardReasoner enhances LLM safety with reasoning-based guardrails, improving performance, explainability, and generalization on various benchmarks.

Large language models underperform humans in open-ended exploration due to prioritizing immediate choices over long-term strategic thinking, but innovative models show promise.