InsightGen uses thematic clustering and graph neighborhood selection to generate diverse, relevant insights for open-ended document-grounded questions and releases the SCOpE-QA dataset of 3000 questions.
Title resolution pending
19 Pith papers cite this work. Polarity classification is still indexing.
19
Pith papers citing it
representative citing papers
Fork-think with confidence identifies forking points via model confidence in a single path before sampling continuations, cutting tokens up to 30% and runtime up to 57% on reasoning benchmarks while matching or exceeding parallel thinking performance.
Finite-answer projections of continuation probabilities stabilize before the answer is parseable, showing 17-31 token mean lead in delayed-verdict tasks with Qwen3-4B-Instruct.
LLMs show significant biases in conflict event classification, with open-weight models exhibiting false illegitimation and adapted models showing actor bias and lexical sensitivity, making them unsuitable for unsupervised deployment.
LLMs recover only ~20% of explicit pragmatic shifts under implicit cultural cues across five languages, responding mainly to linguistic structure rather than cultural associations as shown by Hindi-Urdu controls.
Cross-modal agreement between chain-of-thought and program-of-thought reasoning enables self-consistency with only two LLM samples, reducing sampling cost by 9.3x while improving accuracy.
Evaluator-specific demonstrations with retrospective reasoning improve LLM simulation of individual preference judges by up to 9.9 points over a non-personalized base judge, while interface telemetry often degrades accuracy.
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 depends on model scale.
iPOE generates and optimizes annotation guidelines from explanations to produce interpretable prompts, reporting up to 39% gains over baselines on four datasets with LLM explanations substituting for human ones.
Macro uses DPO on composite preference pairs to raise validity of multilingual self-generated counterfactual explanations by 12.55% on average over chain-of-thought while preserving minimality.
ExPerT infers query-specific user expertise from semantic text and keystroke dynamics via LLM prompting to adapt response generation, cutting inference error 65.7% and raising satisfaction 17.52% in a 40-participant study.
PV-SQL boosts Text-to-SQL execution accuracy by 5% and valid efficiency by 20.8% on BIRD benchmarks via database probing and rule-based SQL verification while using fewer tokens.
Introduces a modality-switching mechanism for LLMs on spatial reasoning tasks using a trustworthiness and complexity based metric, showing up to 42% performance improvement.
UserGPT introduces a generative LLM framework with a behavior simulation engine, semantization module, and DF-GRPO post-training that scores 0.7325 on tag prediction and 0.7528 on summary generation on HPR-Bench while compressing records by up to 97.9%.
SABA improves LLM performance on detective puzzle benchmarks by recursively fusing information into a base state and using queries to resolve missing premises before concluding.
Chain-of-thought monitorability provides a promising but fragile method for AI safety oversight that developers should actively preserve.
Empirical evaluation of quantization effects on eight LLMs across bit widths, showing performance generally declines at lower precision but with model-size-dependent resilience and acceptable accuracy at 2 bits for many cases.
LLMs show susceptibility to presumptions induced by task phrasing in decision tasks like the iterated prisoner's dilemma, mitigated by neutral wording.
citing papers explorer
-
An Answer is just the Start: Related Insight Generation for Open-Ended Document-Grounded QA
InsightGen uses thematic clustering and graph neighborhood selection to generate diverse, relevant insights for open-ended document-grounded questions and releases the SCOpE-QA dataset of 3000 questions.
-
Fork-Think with Confidence
Fork-think with confidence identifies forking points via model confidence in a single path before sampling continuations, cutting tokens up to 30% and runtime up to 57% on reasoning benchmarks while matching or exceeding parallel thinking performance.
-
When Does a Language Model Commit? A Finite-Answer Theory of Pre-Verbalization Commitment
Finite-answer projections of continuation probabilities stabilize before the answer is parseable, showing 17-31 token mean lead in delayed-verdict tasks with Qwen3-4B-Instruct.
-
Are LLMs Ready for Conflict Monitoring? Empirical Evidence from West Africa
LLMs show significant biases in conflict event classification, with open-weight models exhibiting false illegitimation and adapted models showing actor bias and lexical sensitivity, making them unsuitable for unsupervised deployment.
-
Do LLMs Use Cultural Knowledge Without Being Told? A Multilingual Evaluation of Implicit Pragmatic Adaptation
LLMs recover only ~20% of explicit pragmatic shifts under implicit cultural cues across five languages, responding mainly to linguistic structure rather than cultural associations as shown by Hindi-Urdu controls.
-
Self-Consistency from Only Two Samples: CoT-PoT Ensembling for Efficient LLM Reasoning
Cross-modal agreement between chain-of-thought and program-of-thought reasoning enables self-consistency with only two LLM samples, reducing sampling cost by 9.3x while improving accuracy.
-
PERSONAJUDGE: Simulating Individual Human Preference Judgments with Evaluator-Specific Demonstration Data
Evaluator-specific demonstrations with retrospective reasoning improve LLM simulation of individual preference judges by up to 9.9 points over a non-personalized base judge, while interface telemetry often degrades accuracy.
-
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 depends on model scale.
-
iPOE: Interpretable Prompt Optimization via Explanations
iPOE generates and optimizes annotation guidelines from explanations to produce interpretable prompts, reporting up to 39% gains over baselines on four datasets with LLM explanations substituting for human ones.
-
Macro: Enhancing Multilingual Counterfactual Explanations through Alignment-as-Preference Optimization
Macro uses DPO on composite preference pairs to raise validity of multilingual self-generated counterfactual explanations by 12.55% on average over chain-of-thought while preserving minimality.
-
ExPerT: Personalizing LLM Responses to Users' Domain Expertise via Query-Wise Semantic and Keystroke Behavioral Cues
ExPerT infers query-specific user expertise from semantic text and keystroke dynamics via LLM prompting to adapt response generation, cutting inference error 65.7% and raising satisfaction 17.52% in a 40-participant study.
-
PV-SQL: Synergizing Database Probing and Rule-based Verification for Text-to-SQL Agents
PV-SQL boosts Text-to-SQL execution accuracy by 5% and valid efficiency by 20.8% on BIRD benchmarks via database probing and rule-based SQL verification while using fewer tokens.
-
Spatial Reasoning via Modality Switching Between Language and Symbolic Representation
Introduces a modality-switching mechanism for LLMs on spatial reasoning tasks using a trustworthiness and complexity based metric, showing up to 42% performance improvement.
-
UserGPT Technical Report
UserGPT introduces a generative LLM framework with a behavior simulation engine, semantization module, and DF-GRPO post-training that scores 0.7325 on tag prediction and 0.7528 on summary generation on HPR-Bench while compressing records by up to 97.9%.
-
Self-Awareness before Action: Mitigating Logical Inertia via Proactive Cognitive Awareness
SABA improves LLM performance on detective puzzle benchmarks by recursively fusing information into a base state and using queries to resolve missing premises before concluding.
-
Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety
Chain-of-thought monitorability provides a promising but fragile method for AI safety oversight that developers should actively preserve.
-
K-Quantization and its Impact on Output Performance
Empirical evaluation of quantization effects on eight LLMs across bit widths, showing performance generally declines at lower precision but with model-size-dependent resilience and acceptable accuracy at 2 bits for many cases.
-
Impact of Task Phrasing on Presumptions in Large Language Models
LLMs show susceptibility to presumptions induced by task phrasing in decision tasks like the iterated prisoner's dilemma, mitigated by neutral wording.
- How Many Different Outputs Can a Transformer Generate?