paper-with-me

Papers

SPOT: Sparse Probing and Outcome Calibration for On-Policy Distillation

2026-08-05 · Zikun Qu, Min Zhang, Mingze Kong, Zhiwei Shang, Yikun Ban, Shuang Qiu, Zhongxiang Dai hf

On-policy distillation (OPD) provides dense teacher supervision on student-generated trajectories, but standard reverse-KL training can assign insufficient probability to other plausible continuations. Teacher entropy alone does not reveal whether uncertainty is concentrated among a few plausible next tokens or dispersed over a long probability tail, nor whether the student already represents those candidates well. Moreover, local teacher probabilities may not predict downstream success. We introduce Sparse Probing and Outcome-calibrated Targets OPD (SPOT), which addresses two coupled decisions, where to probe and what to distill, through an acquisition--exploration--exploitation procedure. During acquisition, a position-level score combines normalized teacher entropy, the probability mass captured by a small top-$k$ candidate set, and student--teacher mismatch to allocate a limited probing budget. During exploration, SPOT evaluates teacher-proposed candidates through verifier-scored student continuations. During exploitation, these outcomes produce a closed-form, KL-regularized target that favors candidates with better downstream outcomes while remaining anchored to the teacher distribution. Extensive experiments across multiple student models and reasoning benchmarks demonstrate the effectiveness of SPOT in improving reasoning performance while balancing solution quality and coverage.

📄 PDF Abstract BibTeX arXiv:2608.04419

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

DESPOT: Online POMDP Planning with Regularization

2016-09-12 · NeurIPS 2013 12 · Nan Ye, Adhiraj Somani, David Hsu, Wee Sun Lee

The partially observable Markov decision process (POMDP) provides a principled general framework for planning under uncertainty, but solving POMDPs optimally is computationally intractable, due to the "curse of dimension…

Autonomous Driving

The Multipath Blind Spot: $K$-Agnostic Robust Calibration for Sparse-Anchor Metric Depth from Frozen Foundations

2026-07-05 · Sohag Roy, Rajesh Misra, Swami Shastravidyananda, Tamal Maharaj arxiv

Monocular depth foundations predict domain-general relative depth but lack absolute scale; a handful of sparse metric anchors from a range sensor can calibrate them to metric depth, an attractive alternative to metric-su…

Multi-Turn Reinforcement Learning for Tool-Calling Agents with Iterative Reward Calibration

2026-04-03 · Wachiravit Modecrua, Krittanon Kaewtawee, Krittin Pachtrachai, Touchapon Kraisingkorn arxiv

Training tool-calling agents with reinforcement learning on multi-turn tasks remains challenging due to sparse outcome rewards and difficult credit assignment across conversation turns. We present the first application o…

Reinforcement Learning

SPOTR: Spatio-temporal Pooling One-Token Reconstruction for Universal Physiological Signal Self-supervised Learning

2026-06-20 · Yiyu Gui, Mingzhi Chen, Yuesheng Zhu, Guibo Luo 외 arxiv

Physiological signals such as EEG, ECG, and PPG are widely used in clinical monitoring. Recent self-supervised learning (SSL) methods offer an attractive way to leverage unlabeled recordings, yet they still fall short in…

Self-Supervised Learning

Near-Miss: Latent Policy Failure Detection in Agentic Workflows

2026-03-31 · Ella Rabinovich, David Boaz, Naama Zwerdling, Ateret Anaby-Tavor arxiv

Agentic systems for business process automation often require compliance with policies governing conditional updates to the system state. Evaluation of policy adherence in LLM-based agentic workflows is typically perform…