Conformal language modeling samples from posterior approximations conditioned on high-scoring regions to achieve risk control with higher utility than post-hoc filtering in open-ended text generation.
Hal- luHard: A Hard Multi-Turn Hallucination Benchmark
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
years
2026 6representative citing papers
RankJudge creates paired multi-turn conversations with isolated single-turn flaws to generate unambiguous benchmarks for LLM-as-a-judge systems across ML, biomedicine, and finance domains.
Strong medical-model task scores do not guarantee process stability: omission, contradiction, and delay mainly break contradiction detection, diagnosis updating, and hallucination self-correction while final evidence grounding can look fine.
Scoping review of 134 studies on LLM-as-a-Judge in healthcare finds concentration in clinical decision support and NLP, frequent use of OpenAI models with prompt engineering, and moderate-to-strong human alignment where validated.
Empirical tracing across model families shows verification precedes and outlasts generation for facts, with updates producing simultaneous verification of old and new answers.
Empirical study across multiple benchmarks finds the link between uncertainty estimators and LLM hallucinations is highly variable and often weak.
citing papers explorer
-
Conformal Language Modeling via Posterior Sampling
Conformal language modeling samples from posterior approximations conditioned on high-scoring regions to achieve risk control with higher utility than post-hoc filtering in open-ended text generation.
-
RankJudge: A Multi-Turn LLM-as-a-Judge Synthetic Benchmark Generator
RankJudge creates paired multi-turn conversations with isolated single-turn flaws to generate unambiguous benchmarks for LLM-as-a-judge systems across ML, biomedicine, and finance domains.
-
MedBench v5: A Dynamic, Process-Oriented, and Hallucination-Aware Benchmark for Clinical Multimodal Models
Strong medical-model task scores do not guarantee process stability: omission, contradiction, and delay mainly break contradiction detection, diagnosis updating, and hallucination self-correction while final evidence grounding can look fine.
-
LLM-as-a-Judge in Healthcare: A Scoping Analysis of Applications, Methods, and Human Alignment
Scoping review of 134 studies on LLM-as-a-Judge in healthcare finds concentration in clinical decision support and NLP, frequent use of OpenAI models with prompt engineering, and moderate-to-strong human alignment where validated.
-
The Future of Facts: Tracing the Factual Generation-Verification Gap
Empirical tracing across model families shows verification precedes and outlasts generation for facts, with updates producing simultaneous verification of old and new answers.
-
Evaluating the Relevance of Uncertainty Estimators for LLM Hallucination
Empirical study across multiple benchmarks finds the link between uncertainty estimators and LLM hallucinations is highly variable and often weak.