Applying STP at consecutive semantic reasoning steps achieves 168x more accurate multi-step latent prediction on ProcessBench than frozen baselines, with trajectories forming smooth curves best captured by non-linear predictors.
Latent space chain-of-embedding en- ables output-free llm self-evaluation
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
representative citing papers
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.
Consensus Entropy measures inter-VLM output agreement to verify OCR reliability and enable self-improving ensembles, yielding 42.1% F1 gains over single-model judging.
DyCo-RL improves four RLVR algorithms on seven visual and math reasoning benchmarks by assigning tokens visual or text roles via Fisher-Rao geodesic distance on attention and reweighting advantages by role-alignment score.
A two-stage probe (hidden-state estimate plus attention-based correction) yields per-step GRPO rewards that survive prefix contamination and beat external-judge and tree-search rewards in the reported benchmarks.
SIVR detects LLM hallucinations by learning from token-wise and layer-wise variance patterns in internal hidden states, outperforming baselines with better generalization and less training data.
citing papers explorer
-
Semantic Step Prediction: Multi-Step Latent Forecasting in LLM Reasoning Trajectories via Step Sampling
Applying STP at consecutive semantic reasoning steps achieves 168x more accurate multi-step latent prediction on ProcessBench than frozen baselines, with trajectories forming smooth curves best captured by non-linear predictors.
-
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.
-
Consensus Entropy: Harnessing Multi-VLM Agreement for Self-Verifying and Self-Improving OCR
Consensus Entropy measures inter-VLM output agreement to verify OCR reliability and enable self-improving ensembles, yielding 42.1% F1 gains over single-model judging.
-
DyCo-RL: Dynamic Cross-Modal Coordination for Visual Reasoning
DyCo-RL improves four RLVR algorithms on seven visual and math reasoning benchmarks by assigning tokens visual or text roles via Fisher-Rao geodesic distance on attention and reweighting advantages by role-alignment score.
-
PAIR: Prefix-Aware Internal Reward Model for Multi-Turn Agent Optimization
A two-stage probe (hidden-state estimate plus attention-based correction) yields per-step GRPO rewards that survive prefix contamination and beat external-judge and tree-search rewards in the reported benchmarks.
-
Learning Uncertainty from Sequential Internal Dispersion in Large Language Models
SIVR detects LLM hallucinations by learning from token-wise and layer-wise variance patterns in internal hidden states, outperforming baselines with better generalization and less training data.