Flow models reach 99.2% Sudoku accuracy in 7 passes and 96.1% on out-of-distribution Sudoku-Extreme by selecting dynamically stable candidates and training with self-conditioning plus DPO to avoid failed outputs.
hub
Scalable best-of-n selection for large language models via self-certainty.arXiv preprint arXiv:2502.18581
25 Pith papers cite this work. Polarity classification is still indexing.
25
Pith papers citing it
hub tools
citation-role summary
background 3
other 1
citation-polarity summary
representative citing papers
Uncertainty trace profiles from LM reasoning traces predict correct final answers with AUROC up to 0.807 and enable early error detection using only initial tokens.
Joint Consistency casts test-time aggregation as Ising-type energy minimization with pairwise LLM-judge interactions, subsuming voting methods and outperforming baselines across reasoning tasks.
Distinct Leaf Enumeration (DLE) replaces stochastic self-consistency sampling with deterministic traversal of a truncated decoding tree to enumerate distinct leaves, increasing coverage and reducing redundant computation while improving performance on math, coding, and reasoning benchmarks.
Parallel thinking in LLMs suffers from overscaling where fixed global budgets waste samples; LanBo predicts per-sample budgets from latent states to raise utilization without hurting accuracy.
Explorer-Definer and Reflective Orchestrator harnesses raise DeepSeek V3.2 from 15.5% to 67.25% pass@2 on ARC-AGI-1 public eval at $0.25–$0.62 per task without ARC-specific training.
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.
CQC-RAG improves RAG factuality by generating diverse equivalent queries, building query-specific contexts, and selecting answers via cross-query confidence stability, with reported gains of +4.76 pp EM on TriviaQA and +9.12 pp EM on MuSiQue.
LLMs display widespread rational value risk in reasoning that value alignment reduces but does not remove, with risk sensitive to inference strategy and showing diminishing returns from longer reasoning.
Cross-model semantic disagreement adds an epistemic uncertainty term that improves total uncertainty estimation over self-consistency alone, helping flag confident errors in LLMs.
TTSP samples and filters multiple zoom-in exploration traces and iteratively consolidates validated observations into an Evidence Ledger, improving fine-grained multimodal reasoning on V* Bench, HR-Bench, TreeBench, and MME-RealWorld-Lite.
SelfGrader grades query harmfulness from anchored numerical-token logits plus PAC-guided dual-perspective ICL, claiming strong jailbreak detection with low FPR and latency.
Self-distillation turns pretrained autoregressive LMs into multi-token predictors that decode over 3x faster with under 5% accuracy drop on GSM8K.
ETS performs training-free RL alignment for language models by energy-guided test-time scaling with Monte Carlo energy estimation and importance sampling acceleration.
UCAS refines RLVR advantage signals with a logit-space self-confidence proxy for response-level modulation and asymmetric token-level penalties based on raw logit certainty to boost exploration and reduce entropy collapse.
C3RL is a new RL algorithm combining correctness, calibration, and reference accuracy rewards to improve LLM confidence calibration, enabling CAS to outperform majority voting with up to 12.33x lower inference cost.
CAT uses intrinsic confidence signals in preference optimization to adapt reasoning length in LRMs, outperforming uniform compression baselines on accuracy across benchmarks.
Higher sequence probability predicts correctness across different answers in a dataset but does not reliably improve accuracy when decoding methods or hyperparameters are changed, nor does it indicate correctness for repeated responses to one prompt.
EP-HUBO treats CoT evidence selection as higher-order unconstrained binary optimization over per-hypothesis pools with quality weights to improve aggregation on legal benchmarks.
Select to Think reframes LLM help as ranking among SLM top-K candidates and distills the ranking ability back into the SLM for improved single-pass reasoning.
RDS computes total L1 distance of N sampled embeddings from their empirical centroid on the unit hypersphere to measure semantic variability, with a probability-weighted variant that outperforms nine baselines on hallucination detection across four QA datasets and four LLMs.
TokUR estimates token-level uncertainty via low-rank weight perturbations in LLMs, aggregates signals to correlate with correctness, and uses them to improve reasoning performance on math tasks.
CSD adds content-aware entropy relaxation and a distribution alignment filter to speculative decoding, raising acceptance rates in low-detail image areas while keeping output aligned with the target model.
PivotTrace selects unlabeled data for RLVR by quantifying uncertainty via pivot density from attention dynamics, outperforming full supervision using only 29.3% annotations and converging 2.75 times faster.
citing papers explorer
-
Flow Reasoning Models: Scaling Reasoning Through Iterative Self-Refinement
Flow models reach 99.2% Sudoku accuracy in 7 passes and 96.1% on out-of-distribution Sudoku-Extreme by selecting dynamically stable candidates and training with self-conditioning plus DPO to avoid failed outputs.
-
Tracing Uncertainty in Language Model "Reasoning"
Uncertainty trace profiles from LM reasoning traces predict correct final answers with AUROC up to 0.807 and enable early error detection using only initial tokens.
-
Joint Consistency: A Unified Test-Time Aggregation Framework via Energy Minimization
Joint Consistency casts test-time aggregation as Ising-type energy minimization with pairwise LLM-judge interactions, subsuming voting methods and outperforming baselines across reasoning tasks.
-
Efficient Test-Time Inference via Deterministic Exploration of Truncated Decoding Trees
Distinct Leaf Enumeration (DLE) replaces stochastic self-consistency sampling with deterministic traversal of a truncated decoding tree to enumerate distinct leaves, increasing coverage and reducing redundant computation while improving performance on math, coding, and reasoning benchmarks.
-
On the Overscaling Curse of Parallel Thinking: System Efficacy Contradicts Sample Efficiency
Parallel thinking in LLMs suffers from overscaling where fixed global budgets waste samples; LanBo predicts per-sample budgets from latent states to raise utilization without hurting accuracy.
-
Cost-Effective Agent Harnesses for Abstract Reasoning and Generalization on ARC-AGI-1
Explorer-Definer and Reflective Orchestrator harnesses raise DeepSeek V3.2 from 15.5% to 67.25% pass@2 on ARC-AGI-1 public eval at $0.25–$0.62 per task without ARC-specific training.
-
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.
-
CQC-RAG: Robust Retrieval-Augmented Generation via Cross-Query Consistency
CQC-RAG improves RAG factuality by generating diverse equivalent queries, building query-specific contexts, and selecting answers via cross-query confidence stability, with reported gains of +4.76 pp EM on TriviaQA and +9.12 pp EM on MuSiQue.
-
In LLM Reasoning, there is Irrationality on top of Value Misalignment
LLMs display widespread rational value risk in reasoning that value alignment reduces but does not remove, with risk sensitive to inference strategy and showing diminishing returns from longer reasoning.
-
Complementing Self-Consistency with Cross-Model Disagreement for Uncertainty Quantification
Cross-model semantic disagreement adds an epistemic uncertainty term that improves total uncertainty estimation over self-consistency alone, helping flag confident errors in LLMs.
-
Test-time Scaling over Perception: Resolving the Grounding Paradox in Thinking with Images
TTSP samples and filters multiple zoom-in exploration traces and iteratively consolidates validated observations into an Evidence Ledger, improving fine-grained multimodal reasoning on V* Bench, HR-Bench, TreeBench, and MME-RealWorld-Lite.
-
SelfGrader: LLM Jailbreak Detection via Anchored Token-Level Logits
SelfGrader grades query harmfulness from anchored numerical-token logits plus PAC-guided dual-perspective ICL, claiming strong jailbreak detection with low FPR and latency.
-
Multi-Token Prediction via Self-Distillation
Self-distillation turns pretrained autoregressive LMs into multi-token predictors that decode over 3x faster with under 5% accuracy drop on GSM8K.
-
ETS: Energy-Guided Test-Time Scaling for Training-Free RL Alignment
ETS performs training-free RL alignment for language models by energy-guided test-time scaling with Monte Carlo energy estimation and importance sampling acceleration.
-
Unlocking Exploration in RLVR: Uncertainty-aware Advantage Shaping for Deeper Reasoning
UCAS refines RLVR advantage signals with a logit-space self-confidence proxy for response-level modulation and asymmetric token-level penalties based on raw logit certainty to boost exploration and reduce entropy collapse.
-
Scaling with Confidence: Calibrating Confidence of LLMs for Adaptive Test Time Scaling
C3RL is a new RL algorithm combining correctness, calibration, and reference accuracy rewards to improve LLM confidence calibration, enabling CAS to outperform majority voting with up to 12.33x lower inference cost.
-
CAT: Confidence-Adaptive Thinking for Efficient Reasoning of Large Reasoning Models
CAT uses intrinsic confidence signals in preference optimization to adapt reasoning length in LRMs, outperforming uniform compression baselines on accuracy across benchmarks.
-
When are likely answers right? On Sequence Probability and Correctness in LLMs
Higher sequence probability predicts correctness across different answers in a dataset but does not reliably improve accuracy when decoding methods or hyperparameters are changed, nor does it indicate correctness for repeated responses to one prompt.
-
Quantum-Inspired Trace-Augmented Evidence Selection for Reasoning over Structured Hypothesis Spaces
EP-HUBO treats CoT evidence selection as higher-order unconstrained binary optimization over per-hypothesis pools with quality weights to improve aggregation on legal benchmarks.
-
Select to Think: Unlocking SLM Potential with Local Sufficiency
Select to Think reframes LLM help as ranking among SLM top-K candidates and distills the ranking ability back into the SLM for improved single-pass reasoning.
-
Distance Is All You Need: Radial Dispersion for Uncertainty Estimation in Large Language Models
RDS computes total L1 distance of N sampled embeddings from their empirical centroid on the unit hypersphere to measure semantic variability, with a probability-weighted variant that outperforms nine baselines on hallucination detection across four QA datasets and four LLMs.
-
TokUR: Token-Level Uncertainty Estimation for Large Language Model Reasoning
TokUR estimates token-level uncertainty via low-rank weight perturbations in LLMs, aggregates signals to correlate with correctness, and uses them to improve reasoning performance on math tasks.
-
CSD: Content-aware Speculative Decoding for Efficient Image Generation
CSD adds content-aware entropy relaxation and a distribution alignment filter to speculative decoding, raising acceptance rates in low-detail image areas while keeping output aligned with the target model.
-
Smart Picks in the Dark: Towards Efficient RLVR for Reasoning via Tracing Metacognitive Pivots
PivotTrace selects unlabeled data for RLVR by quantifying uncertainty via pivot density from attention dynamics, outperforming full supervision using only 29.3% annotations and converging 2.75 times faster.
-
A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence
The paper delivers the first systematic review of self-evolving agents, structured around what components evolve, when adaptation occurs, and how it is implemented.