arXiv Computation and Language

Can LLMs Take Retrieved Information with a Grain of Salt?

The paper investigates how large language models (LLMs) handle the certainty of retrieved information, a critical yet underexplored aspect of retrieval-augmented generation. Eight LLMs were evaluated on their ability to adjust responses according to expressed context certainty, revealing systematic shortcomings such as failure to reproduce prior outputs after encountering uncertainty and distorted responses to expressed certainties. The authors propose an interaction strategy—combining prior reminders, context-certainty recalibration, and context simplification—that reduces obedience errors by 25% on average without altering model weights, demonstrating that interaction design can enhance LLM reliability.

arXiv Machine Learning
Sep 3

Prompting the Unknown: Understanding Response Uncertainty in Large Language Models

The paper introduces a prompt-response concept model that links the amount of task-relevant information in a prompt to the uncertainty of responses generated by large language models (LLMs). It identifies four sources of response uncertainty—prompt underspecification, model quality, task variability, and semantic redundancy—and demonstrates that uncertainty decreases as prompt informativeness or model quality increases, analogous to epistemic uncertainty in probabilistic models. Experiments on real-world datasets confirm the theoretical predictions and validate the model.

By Ze Yu Zhang, Arun Verma, Finale Doshi-Velez, Bryan Kian Hsiang Low
arXiv AI
Aug 20

Same Facts, Different Updates: Inference Setup Shapes LLM Behavior in Medical Allocation

The paper investigates how the inference setup of large language models (LLMs) influences their behavior in a medical resource‑allocation scenario. By comparing paired‑context and independent‑inference experiments, the authors show that adding a single contrasting patient sentence can shift the model’s probability assignments in opposite directions across most tested models. Additional experiments varying scenario attributes further demonstrate that patient information can have context‑dependent effects on LLM outputs.

By Spencer Gibson, Tyler Crosse, Magnus Saebo, Achyutha Menon, Eyon Jang, Diogo Cruz
arXiv AI
Sep 17

From 'May' to 'Is': Certainty Distortion in Language Model Rewriting

The study examines how language models (LMs) alter the expressed certainty of statements when rewriting text, a process termed certainty distortion. Using an LM‑based metric aligned with human judgments, the authors find that up to 75% of LM outputs exhibit such distortion, with most models more likely to inflate certainty than reduce it. Repeated paraphrasing can amplify this effect, especially in medical contexts, and while prompt interventions help, they do not fully eliminate the bias.

By Catarina G Belem, Shang Wu, Hongyu Yao, Mark Steyvers, Sameer Singh, Padhraic Smyth
arXiv AI
Nov 6

Evaluating Token Probabilities as Confidence Estimators in Human Behavior Simulations: A Case Study in Luxembourgish Next Like Prediction

The paper investigates whether token probabilities from open‑weight large language models can serve as confidence estimates in a Next‑Like Prediction task on the Luxembourgish news platform RTL.lu. Using in‑context learning, the models were found to be overconfident and outperformed by a simple statistical baseline in both accuracy and calibration. Fine‑tuning the models reduced expected calibration error dramatically while also improving accuracy, highlighting that confidence estimation should be evaluated alongside accuracy for valid LLM‑based simulations.

By Nils Schwager, Simon M\"unker, Alistair Plum, Achim Rettinger