AI Paper Atlas

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

OpenEstimate: Evaluating LLMs on Probabilistic Estimation with Real-World Data

Alana Marzoev, Jillian Ross, Jacob Andreas

ICLR 2026 · Large language models › Language, safety and interpretability · approach: Uncertainty quantification for LLMs

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
combine LLM prior with 5 samples instead of uninformative prioro4-mini · OpenEstimate Glassdoorwin rate (%): 37.0% prior win -> 71.4% posterior winImproved
increase baseline sample size from 5 to 30o4-mini prior · OpenEstimate Glassdoorwin rate of LLM prior vs baseline (%): 37.0% -> 8.7%Hurt
increase baseline sample size from 5 to 30o4-mini prior · OpenEstimate NHANESwin rate of LLM prior vs baseline (%): 74.3% -> 37.8%Hurt
increase baseline sample size from 5 to 30o4-mini prior · OpenEstimate Pitchbookwin rate of LLM prior vs baseline (%): 50.8% -> 50.8%No clear effect
increase sample size from 5 to 30 for LLM posterioro4-mini posterior · OpenEstimate Pitchbookwin rate (%): 69.6% -> 81.6%Improved
increase sample size from 5 to 30 for LLM posterioro4-mini posterior · OpenEstimate NHANESwin rate (%): 70.4% -> 50.4%Hurt
use reasoning model o4-mini instead of Llama-3-8BLLM prior · OpenEstimate NHANESCRPS ratio: 19.17 -> 1.17Improved
use reasoning model o4-mini instead of GPT-4oLLM prior · OpenEstimate GlassdoorCRPS ratio: 3.31 -> 2.42Improved
use o3-mini instead of GPT-4oLLM prior · OpenEstimate NHANESCRPS ratio: 1.86 -> 1.35Improved
vary temperature or reasoning effort (ablation)o4-mini and gpt-4o · OpenEstimate (all domains)error ratio, CRPSNo clear effect
vary system prompt (ablation)o4-mini and gpt-4o · OpenEstimate (all domains)error ratio, CRPSNo clear effect
change elicitation protocol (direct, quantile, mean-variance) (ablation)o4-mini and gpt-4o · OpenEstimate (all domains)error ratio, CRPSNo clear effect

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