Tracked topic
…Yang Tian , , , , , , , Abstract Privileged Bayesian Self-Distillation enables fine-grained credit assignment in long-horizon tasks by converting sparse outcome rewards into calibrated turn-level signals through Bayesian evidence scoring and autoregressive…
…Unlike existing approaches that require calibrated, posed RGB-D input, RADIO-ViPE operates directly on raw monocular RGB video streams, requiring no prior camera intrinsics , depth sensors , or pose initialization . The system…
…Scaling test-time compute is a promising way to improve these agents, but current approaches can fail, because correct answers are often sparse and score-based selection depends on model calibration. We…
…We evaluate four deep research agents using sparse and dense multilingual retrievers , measuring answer accuracy, evidence recall , search behavior, calibration , citation fidelity , and oracle retrieval . Results reveal substantial degradation when evidence is…
…RPT further supports confidence-aware optimization by using calibration signals in diagnostic feedback and final prompt selection. Across three reasoning tasks, RPT improves over initial prompts by up to 12.9 points…
…Songtao Wei , , , , , , , , , Abstract LEAD is a method that dynamically adapts reasoning efficiency during training by using online calibration of correctness-efficiency trade-offs and adaptive problem-specific length targets to improve mathematical…
…Sashank Pisupati , Henry Broomfield , Eujeong Choi , , , , , Abstract A large-scale dataset of human-metaevaluations of LLM critiques for agentic tasks is introduced to improve the calibration and reliability of automated evaluation methods…
…ClawGym-Bench — 200 instances, auto-filtered & human-LLM reviewed, calibrated for trust Resources dropping soon https://github.com/ClawGym the dual route data synthesis, fusing persona-driven top-down intents with skill…
…Lightweight retention gates assign utility scores to cached KV entries, and a shared final scoring projection calibrates these scores across all layers and heads. This enables a single global eviction policy in…
…We introduce KVarN , a calibration-free KV-cache quantizer that applies a Hadamard rotation followed by a dual-scaling variance normalization across both axes of the K and V matrices. We find…