Disruption profiles are higher quality explanations than metadata-only descriptions

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Disruption profilesmethod0.793
  Mechanistic explanation outputs from EVEE showing how variants affect gene function, scored 3.8/5 for explanation quality.
• disruption profileconcept0.765
  The output explanation format of Evee, quantifying how a variant disrupts different genomic features.
• Future models with substantially increased capabilities will exhibit alignment faking that is more consistent, robust, and harder to detecthypothesis0.731
  Extrapolation from scale-emergence finding to future risk
• Probe-based data attribution effectively reduces harmful behaviors via data interventionsclaim0.724
  Authors' central interpretive assertion that their method meaningfully mitigates unwanted behaviors.
• Synthetic document fine-tuning avoids artificially strengthening the evaluation-deployment representational direction compared to direct demonstration fine-tuningclaim0.717
  Methodological justification for using SDF over direct demonstrations to train a realistic model organism.
• Probe-based data attribution bridges interpretability work (probes/activations) with data-centric alignment.claim0.716
  Conceptual framing: integrates mechanistic interpretability tools with alignment-focused data curation.
• Larger hidden representations create more random structure that DAS can search through, allowing manipulation of counterfactual behavior even in randomly initialized networkshypothesis0.716
  Tested in Section 4.4 calibration experiment; confirmed by findings.
• NLA explanations confabulate false specifics but maintain thematic fidelity; claims repeated across tokens more likely true than isolated claims.claim0.715
  Core limitation and usage heuristic: read NLAs for themes rather than individual factual claims; cross-check with original context.