When Teachers Mislead: Spurious-Signal-Aware On-Policy Distillation
On-Policy distillation (OPD) transfers teacher capabilities by supervising student-sampled trajectories with dense token-level teacher signals. Recent selective OPD methods improve this process by prioritizing signals that are confident, informative, or learnable. However, the assumptions overlook a fundamental failure mode of language models: their token-level judgments can be driven by input-agnostic language priors, formatting conventions, or stereotyped reasoning templates rather than task-s
Record details
Published: 3 August 2026
Source: HuggingFace Daily Papers
Category: Research
Topics: Regulation · Children & education
Retrieved: 6 August 2026
Related evidence
These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.
Any-OPD: Heterogeneous On-Policy Distillation for Flow-Matching Models via Representation-Space Bridging
HuggingFace Daily Papers · 3 August 2026
DAPD: Dual-Anchored Policy Distillation
arXiv · 3 August 2026
Policy Fragmentation or Institutional Alignment? Institutional Governance of AI in Universities and Business Schools
arXiv · 4 August 2026
Beyond Compliance: A Proposed Framework for Ethical Governance of Student Data in Learning Analytics
arXiv · 4 August 2026
SPOT: Sparse Probing and Outcome Calibration for On-Policy Distillation
HuggingFace Daily Papers · 4 August 2026
When Privileged Guidance Misaligns: State-Matched Routing and Contextualized Self-Distillation for Multi-Turn Agents
HuggingFace Daily Papers · 4 August 2026
How to cite this record
ethics.ai (3 August 2026), “When Teachers Mislead: Spurious-Signal-Aware On-Policy Distillation,” evidence record 16602, https://ethics.ai/record/16602 (originally published by HuggingFace Daily Papers).
Use and limitations
This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.