Simple lexical constraints trigger 14-48% comprehensiveness loss in instruction-tuned LLMs via planning failures, with two-pass generation and probes recovering much of the deficit while base models remain unaffected.
Important:Brevity alone does not constitute conciseness
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
-
One Token Away from Collapse: The Fragility of Instruction-Tuned Helpfulness
Simple lexical constraints trigger 14-48% comprehensiveness loss in instruction-tuned LLMs via planning failures, with two-pass generation and probes recovering much of the deficit while base models remain unaffected.