Bad contexts in LLM conversations cause error repetition, mode collapse, and opinion flipping with 38-40% performance drops that worsen over turns, mitigated by RLVR trained with synthetic errors.
" " Human solution to Problem 2
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Pigeonholing: Bad prompts hurt models to collapse and make mistakes
Bad contexts in LLM conversations cause error repetition, mode collapse, and opinion flipping with 38-40% performance drops that worsen over turns, mitigated by RLVR trained with synthetic errors.