Hallucinations in LVLMs largely arise from textual priors in prompts, and can be reduced by fine-tuning with preference optimization on grounded vs. hallucinated response pairs.
Disentangling length bias in preference learning via response-conditioned modeling
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
verdicts
UNVERDICTED 3representative citing papers
Introduces secondary risks as a new class of LLM failures from benign prompts, defines two primitives, proposes SecLens search framework, and releases SecRiskBench showing risks are widespread across 16 models.
EqLen reframes length bias in sequence-level RL as a comparison-unit construction problem and builds equal-length training segments via dual-track generation, prefix inheritance, and segment masking.
citing papers explorer
-
When Prompts Override Vision: Prompt-Induced Hallucinations in LVLMs
Hallucinations in LVLMs largely arise from textual priors in prompts, and can be reduced by fine-tuning with preference optimization on grounded vs. hallucinated response pairs.
-
Exploring the Secondary Risks of Large Language Models
Introduces secondary risks as a new class of LLM failures from benign prompts, defines two primitives, proposes SecLens search framework, and releases SecRiskBench showing risks are widespread across 16 models.
-
Rethinking the Comparison Unit in Sequence-Level Reinforcement Learning: An Equal-Length Paired Training Framework from Loss Correction to Sample Construction
EqLen reframes length bias in sequence-level RL as a comparison-unit construction problem and builds equal-length training segments via dual-track generation, prefix inheritance, and segment masking.