AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Trained on Tokens, Calibrated on Concepts: The Emergence of Semantic Calibration in LLMs

Preetum Nakkiran, Arwen Bradley, Adam Golinski, Eugene Ndiaye, Michael Kirchhof, Sinead Williamson

ICLR 2026 · Large language models › Language, safety and interpretability · approach: Statistical frameworks for LLM evaluation · 14 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
none (standard pretraining, immediate answer)base LLMs (Qwen, Llama, Mistral, Gemini) · GSM8K, OpenMathInstruct-2, TriviaQA, SimpleQAsemantic calibration errorImproved
increase model size from small to large (ablation)base LLMs · 4 QA datasets, concise/sentencesemantic calibration errorNo clear effect
change response style from concise to sentence (ablation)base LLMs · 4 QA datasetssemantic calibration errorNo clear effect
post-train with instruction DPO (ablation)Mistral-7B-v0.1 · semantic calibration (sentence style)calibration errorHurt
post-train with instruction SFT only (ablation)Mistral-7B-v0.1 · semantic calibration (sentence style)calibration errorNo clear effect
post-train with RLHF/DPO/RLVR instruct models (ablation)instruct LLMs · 4 QA datasetssemantic calibration errorHurt
allow chain-of-thought reasoning before answering (ablation)base LLMs · GSM8K, OpenMathInstruct-2semantic calibration errorHurt
increase model capability (accuracy) (ablation)base LLMs in predicted-calibrated settings · 4 QA datasetscalibration errorNo clear effect
lower LoRA KL gap in predicting own semantic distribution (rank-8 LoRA) (ablation)Qwen2.5 base 0.5B-14B · GSM8Kcalibration errorImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy