Fidelity Is Not Safety: Gently-Compressed LLMs Pass Every Data-Free Quality Guard Yet Invent Procedure Steps in Agentic Execution
Ranking
Overall
90
Content
100
Popularity
66
Observed public metrics from 1 member.
Merged summary
TL;DR - Gently compressed LLMs can pass standard quality and fidelity checks yet invent procedural steps during agentic execution. A data-free compression-error screen identifies risky low-rank builds before deployment.
- SVD truncation caused invented SOP steps across three model families; perplexity-matched magnitude pruning did not.
- Perplexity, MMLU, and representation-based fidelity tests failed to predict this behavior.
- Risk correlated with compression-error coherence multiplied by error rate, not overall damage magnitude.
- Fixed thresholds for coherent-error fraction and error rate flagged failing builds across architectures.
Sources (1)
Fidelity Is Not Safety: Gently-Compressed LLMs Pass Every Data-Free Quality Guard Yet Invent Procedure Steps in Agentic Execution
Public signals
Hugging Face upvotes 1
TL;DR - Gently compressed LLMs can pass standard quality and fidelity checks yet invent procedural steps during agentic execution. A data-free compression-error screen identifies risky low-rank builds before deployment.
- SVD truncation caused invented SOP steps across three model families; perplexity-matched magnitude pruning did not.
- Perplexity, MMLU, and representation-based fidelity tests failed to predict this behavior.
- Risk correlated with compression-error coherence multiplied by error rate, not overall damage magnitude.
- Fixed thresholds for coherent-error fraction and error rate flagged failing builds across architectures.