Large Language Models

Structured pruning of self-distilled reasoning traces reduces token use while largely preserving accuracy in low-data fine-tuning.

Sep 1, 2026 • 1 min read

A benchmark and analysis of when tool-using LLM agents should stop and abstain rather than continue acting.

May 1, 2026 • 1 min read

A modular abstention framework for reliable expert LLMs that enables selective abstention from uncertain questions.

Jul 1, 2025 • 1 min read

Exploring psychological insights to address overconfidence in LLMs by comparing with human confidence patterns.

May 1, 2025 • 1 min read

Automatic prediction of compute-optimal data composition for efficient LLM training.

May 1, 2025 • 1 min read

A comprehensive survey of abstention mechanisms in large language models, covering theory, implementation, and evaluation.

Feb 1, 2025 • 1 min read

Behavioral analysis and mitigation strategies for overconfidence in large language models.

Dec 1, 2024 • 1 min read

Characterizing LLM abstention behavior in science QA with context perturbations.

Oct 1, 2024 • 1 min read