REVIEW 14 cited by
Generating with Confidence: Uncertainty Quantification for Black-box Large Language Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Large language models (LLMs) specializing in natural language generation (NLG) have recently started exhibiting promising capabilities across a variety of domains. However, gauging the trustworthiness of responses generated by LLMs remains an open challenge, with limited research on uncertainty quantification (UQ) for NLG. Furthermore, existing literature typically assumes white-box access to language models, which is becoming unrealistic either due to the closed-source nature of the latest LLMs or computational constraints. In this work, we investigate UQ in NLG for *black-box* LLMs. We first differentiate *uncertainty* vs *confidence*: the former refers to the ``dispersion'' of the potential predictions for a fixed input, and the latter refers to the confidence on a particular prediction/generation. We then propose and compare several confidence/uncertainty measures, applying them to *selective NLG* where unreliable results could either be ignored or yielded for further assessment. Experiments were carried out with several popular LLMs on question-answering datasets (for evaluation purposes). Results reveal that a simple measure for the semantic dispersion can be a reliable predictor of the quality of LLM responses, providing valuable insights for practitioners on uncertainty management when adopting LLMs. The code to replicate our experiments is available at https://github.com/zlin7/UQ-NLG.
Forward citations
Cited by 14 Pith papers
-
ACS: An interactive framework for conformal selection
ACS provides finite-sample false discovery rate control for interactive, adaptive selection of promising candidates from unlabeled pools.
-
Reliability Scaling Laws for Quantized Large Language Models
Reliability of quantized LLMs peaks nonlinearly at 4-bit precision under fixed total model bits, while accuracy scales monotonically, and quantization can improve robustness to natural perturbations.
-
Estimating Uncertainty from Reasoning: A Large-Scale Study of Multi- and Crosslingual MCQA Performance in LLMs
A large-scale multilingual evaluation of LLM uncertainty estimation methods across 22 languages and 9 models finds that English reasoning closes the UE gap for low-resource languages and that optimal UE method choice ...
-
Check Yourself Before You Wreck Yourself: Selectively Quitting Improves LLM Agent Safety
Explicitly instructing LLM agents to quit uncertain or risky tasks improves their safety scores by 0.39 on average in ToolEmu simulations, with a negligible average helpfulness decrease of 0.03.
-
Unsupervised Hallucination Detection by Inspecting Reasoning Processes
IRIS detects LLM hallucinations by training a lightweight probe on hidden states elicited during the model's own step-by-step verification, using the model's verbalized confidence as soft pseudolabels.
-
Surfacing Variations to Calibrate Perceived Reliability of MLLM-generated Image Descriptions
Surfacing variations across multiple MLLM image descriptions increases blind and low vision users' detection of unreliable claims and reduces their over-trust in a single AI description.
-
The "I Don't Know" Filter: Enhancing Agentic Reliability in Function Calling
A multi-sample uncertainty classifier filter improves agent reliability (IDKS) by abstaining on uncertain function calls across open SLMs and BFCL-style benchmarks.
-
Beyond Self-Knowledge: Propagating Uncertainty Across Reasoning and Retrieval in LLMs
A verbalized-confidence router improves QA F1 by 0.016 over always retrieval and beats a route-count-matched random control in 17/18 cells, but the probe adds 28.2% tokens.
-
Overconfidence in LLM-as-a-Judge: Diagnosis and Confidence-Driven Solution
LLM-as-a-Judge systems report confidence that overstates their accuracy, and the paper's TH-Score plus LLM-as-a-Fuser improves calibration.
-
Cleanse: Uncertainty Estimation Approach Using Clustering-based Semantic Consistency in LLMs
Cleanse detects hallucinated LLM answers by computing the share of hidden-embedding cosine similarity that falls inside semantic clusters, and it beats several baselines in AUROC across four models and two QA benchmarks.
-
Learning the Pareto Frontier of Predictive Models under Distribution Shift
Frontier Learning — concatenating pretrained models' representations and predictions and fitting a regularized linear/logistic head on target labels — matches or improves on the best individual zero-shot, fine-tuned, ...
-
Trustworthy Agents for Electronic Health Records through Confidence Estimation
TrustEHRAgent, an EHR assistant with step-by-step confidence checks, scores 44% and 25% accuracy on MIMIC-III and eICU when only answers it is 70% confident in are counted, while baseline methods score 0%.
-
Confidence Estimation for Text-to-SQL in Large Language Models
Consistency-based methods are the most reliable confidence signal for text-to-SQL in black-box LLMs, and executing queries against a database adds a useful correctness signal.
-
Building Task Bots with Self-learning for Enhanced Adaptability, Extensibility, and Factuality
A thesis that combines self-learning from dialog logs, schema-guided prompting, and self-aligned factuality to build task bots with minimal human intervention.
Discussion (0). Sign in to comment.