VPD frames language feedback learning as variational EM so the teacher policy refines itself via trust-region updates on outcomes while the student learns dense token distributions on its own rollouts, outperforming fixed-teacher baselines on reasoning and code tasks.
Language models can learn from verbal feedback without scalar rewards.arXiv preprint arXiv:2509.22638, 2025
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
Neighbor-Consistency Belief (NCB) measures LLM belief robustness across conceptual neighborhoods, revealing that high-NCB facts resist contextual interference better, and Structure-Aware Training reduces brittleness by about 30%.
citing papers explorer
-
Learning from Language Feedback via Variational Policy Distillation
VPD frames language feedback learning as variational EM so the teacher policy refines itself via trust-region updates on outcomes while the student learns dense token distributions on its own rollouts, outperforming fixed-teacher baselines on reasoning and code tasks.
-
Illusions of Confidence? Diagnosing LLM Truthfulness via Neighborhood Consistency
Neighbor-Consistency Belief (NCB) measures LLM belief robustness across conceptual neighborhoods, revealing that high-NCB facts resist contextual interference better, and Structure-Aware Training reduces brittleness by about 30%.