MiCP is the first conformal prediction method for multi-turn LLM pipelines that allocates per-turn error budgets to enable adaptive stopping with an overall coverage guarantee, shown to reduce turns and cost on RAG and ReAct benchmarks.
Title resolution pending
11 Pith papers cite this work. Polarity classification is still indexing.
11
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
years
2026 11roles
background 2polarities
background 2representative citing papers
Compositional selective specificity (CSS) decomposes generated answers into claims and emits each at the most specific level supported by evidence, raising overcommitment-aware utility from 0.846 to 0.913 on LongFact while retaining 0.938 specificity.
Introduces policy-coupled coverage for conformal prediction in counterfactual decisions and the PC-RACP procedure that achieves higher utility with finite-sample coverage guarantees.
REALISTA generates semantically coherent adversarial prompts via latent-space optimization over input-dependent editing directions, achieving stronger hallucination elicitation than prior realistic attacks on open-source and reasoning LLMs.
PASC converts multi-stage joint coverage into a single scalar conformal problem on the joint max nonconformity score, delivering finite-sample distribution-free guarantees and higher empirical coverage than Bonferroni or independent calibration.
Geometry-calibrated conformal abstention lets language models abstain from uncertain queries with finite-sample guarantees on both participation rate and conditional correctness of answers.
The paper introduces an optimization framework for AI agents to strategically seek support, proving a threshold policy on support value and providing an online algorithm to control missed-support error without distributional assumptions.
LLMs recover coherent, interpretable structures from HTS literature including family-dependent mechanisms and temporal belief evolution via a constructed knowledge graph.
Learning a margin-based confidence ranker for LLM judges improves agreement-target success in cascaded selective evaluation compared to heuristic confidence scores.
IUQ quantifies claim-level uncertainty in long-form LLM generation by combining inter-sample consistency and intra-sample faithfulness through an interrogate-then-respond approach and outperforms baselines on two datasets.
Simple thresholding on an external verifier signal, calibrated by risk control, performs competitively with sequential hypothesis testing monitors on math reasoning and red-teaming datasets.
citing papers explorer
-
Adaptive Stopping for Multi-Turn LLM Reasoning
MiCP is the first conformal prediction method for multi-turn LLM pipelines that allocates per-turn error budgets to enable adaptive stopping with an overall coverage guarantee, shown to reduce turns and cost on RAG and ReAct benchmarks.
-
Answer Only as Precisely as Justified: Calibrated Claim-Level Specificity Control for Agentic Systems
Compositional selective specificity (CSS) decomposes generated answers into claims and emits each at the most specific level supported by evidence, raising overcommitment-aware utility from 0.846 to 0.913 on LongFact while retaining 0.938 specificity.
-
Prediction Sets for Counterfactual Decisions: Coverage, Optimality, and Conformal Prediction
Introduces policy-coupled coverage for conformal prediction in counterfactual decisions and the PC-RACP procedure that achieves higher utility with finite-sample coverage guarantees.
-
REALISTA: Realistic Latent Adversarial Attacks that Elicit LLM Hallucinations
REALISTA generates semantically coherent adversarial prompts via latent-space optimization over input-dependent editing directions, achieving stronger hallucination elicitation than prior realistic attacks on open-source and reasoning LLMs.
-
PASC: Pipeline-Aware Conformal Prediction with Joint Coverage Guarantees for Multi-Stage NLP and LLM Pipelines
PASC converts multi-stage joint coverage into a single scalar conformal problem on the joint max nonconformity score, delivering finite-sample distribution-free guarantees and higher empirical coverage than Bonferroni or independent calibration.
-
Geometry-Calibrated Conformal Abstention for Language Models
Geometry-calibrated conformal abstention lets language models abstain from uncertain queries with finite-sample guarantees on both participation rate and conditional correctness of answers.
-
Strategic Decision Support for AI Agents
The paper introduces an optimization framework for AI agents to strategically seek support, proving a threshold policy on support value and providing an online algorithm to control missed-support error without distributional assumptions.
-
Can LLMs extract scientific consensus? A case study in high-temperature superconductivity
LLMs recover coherent, interpretable structures from HTS literature including family-dependent mechanisms and temporal belief evolution via a constructed knowledge graph.
-
Margin-Adaptive Confidence Ranking for Reliable LLM Judgement
Learning a margin-based confidence ranker for LLM judges improves agreement-target success in cascaded selective evaluation compared to heuristic confidence scores.
-
IUQ: Interrogative Uncertainty Quantification for Long-Form Large Language Model Generation
IUQ quantifies claim-level uncertainty in long-form LLM generation by combining inter-sample consistency and intra-sample faithfulness through an interrogate-then-respond approach and outperforms baselines on two datasets.
-
Online Safety Monitoring for LLMs
Simple thresholding on an external verifier signal, calibrated by risk control, performs competitively with sequential hypothesis testing monitors on math reasoning and red-teaming datasets.