Tracked topic
…On success-inferable tasks, analyzer-based early stopping further prevents late-stage performance regressions and saves 25%-53% of attempts when the success signal is properly calibrated. View arXiv page View PDF…
…Tau calibration — thresholds are currently post-training and corpus-specific. Online adaptation during training feels like the right direction. Any pointers to related work on dynamic threshold learning? All code, training configs…
…Improving Long-Form Generation Factuality via Reasoning Calibration (2026) Too Correct to Learn: Reinforcement Learning on Saturated Reasoning Data (2026) AGPO: Asymmetric Group Policy Optimization for Verifiable Reasoning and Search Ads Relevance…
…Signal-Calibrated On-Policy Distillation Enhancement with Dual-Path Adaptive Weighting (2026) Correlation-Weighted Multi-Reward Optimization for Compositional Generation (2026) Beyond Static Best-of-N: Bayesian List-wise Alignment for LLM…
…Orchestrating Safety-Aware Visual Narratives via Multi-Agent Cognitive Calibration (2026) Please give a thumbs up to this comment if you found it helpful! If you want recommendations for any Paper on…
…Adaptive Verification for MoE Speculative Decoding (2026) Calibrated Speculative Decoding: Frequency-Guided Candidate Selection for Efficient Inference (2026) TAPS: Target-Aware Prefix Tree Selection for Diffusion-Drafted Speculative Decoding (2026) Draft-OPD…
…The Squeeze-Release cycle iterates pruning and minimization with an intermediate release step that re-enables the exact-zero positions inside the compacted tensors as small calibrated noise, turning otherwise wasted capacity…
…true # Each step must have thought + action = freedom confidence_calibration: true # Post-process confidence scores ciritcally high! Congrads! Get Pi as the agent harness next · cc: @ victor A great milestone for Local…
…Replications across scale, family, and rank show the behavioral detector transfers without retuning, while the weight-level detector is calibration-bound to the base model. The attack scales monotonically with rank, and…
…Training-Free Calibration for MoE (2026) Please give a thumbs up to this comment if you found it helpful! If you want recommendations for any Paper on Hugging Face checkout this Space…