AI Paper Atlas

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

Eliciting Numerical Predictive Distributions of LLMs Without Auto-Regression

Julianna Piskorz, Kasia Kobalczyk, Mihaela van der Schaar

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
use magnitude-factorised probe instead of vanilla MLP probe with log-scaling (ablation)regression probe · Synthetic scale 1.0, greedy targetMSE vs LLM statistic: 0.0400 -> 0.0150Improved
use magnitude-factorised probe instead of vanilla MLP probe (no log-scaling) (ablation)regression probe · Synthetic scale 1.0, mean targetMSE vs LLM statistic: 0.0092 -> 0.0061Improved
two-phase training of order and value heads instead of joint training (ablation)magnitude-factorised point probe · Synthetic dataMSEImproved
two-phase training instead of joint training for quantile probe (ablation)quantile probe · Synthetic dataquantile accuracyNo clear effect
concatenate hidden states from multiple layers instead of a single layer (ablation)probe input · Synthetic data, Llama-2-7Bprobe errorImproved
train probe on narrow context length range [10, 20] instead of [3, 40] (ablation)quantile probe · Unseen context lengths, scale 1.0coverage deviationHurt
train probe on real data from all sub-datasets versus holding out sub-datasets (ablation)quantile probe · Real-world data, 90% interval coverageempirical coverage (%): 82.1 -> 88.5Improved
train probe on synthetic data only and test on real data (ablation)quantile probe · Real-world data, 90% interval coverageempirical coverage (%): 88.5 -> 67.7Hurt
use probe in place of 20-25 LLM samples for next-value predictionmagnitude-factorised probe · Synthetic scale 1.0MSE vs ground truth: probe outperforms empirical sampling for N up to 20-25 samplesImproved
use probe instead of autoregressive sampling for one predictionLlama-2-7B · Inference timetime (s): 1.59 -> 0.034Improved
probe predicts greedy output instead of meanmagnitude-factorised probe · Synthetic scale 1.0MSE vs LLM statistic: 0.006 -> 0.015Hurt
probe prediction versus LLM's own accuracy on next valueprobe · Synthetic scale 1.0MSE vs ground truth: 0.0562 vs 0.0555No 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