record state
frontier-ownedReview status
This finding is part of accepted frontier state. Review events, reviewable changes, and proof state explain how it can change.
unreviewed
frontiers / frontier
Finding bundle
back to stateno incoming links yet
record state
frontier-ownedThis finding is part of accepted frontier state. Review events, reviewable changes, and proof state explain how it can change.
finding statement
finding typeNo entity list is declared.
evidence
source-boundcomputational · manual state transition
proof impact
packet context2 reviewable changes and 0 evaluation records are attached to this finding id.
Evidence and conditions
method
manual state transition
evidence type
computational
conditions
Provenance
source title
mechinterp circuit harness
authors
agent:replicator
In gpt2, attention head L4H11 acts as a previous token head (min held-out score 0.9823 across 5 prompts; role also present in distilgpt2).
vs_f1c2a9e52cbd0c39 · manual_curation
outgoing
No outgoing links.
incoming
No incoming links.
events
vev_ad5bb86e78f2880efinding.assertedManual finding added to frontier state
agent:replicator · 2026-05-29
vev_37a6454fcbbc4d40finding.caveatedVerifier caution: L4H11 as GPT-2's canonical previous-token head is strongly confirmed across independent sources (your README, LessWrong, mechanistic interpretability literature). Explicitly listed in mechinterp/README.md as a known example. Score 0.9823 matches expected strength. However, no rep
reviewer:will-blair · 2026-05-29
reviewable changes
vpr_d5ccff74896ee56afinding.addManual finding added to frontier state
applied · agent:replicator · 2026-05-29
vpr_800fd5eafc83dcebfinding.caveatVerifier caution: L4H11 as GPT-2's canonical previous-token head is strongly confirmed across independent sources (your README, LessWrong, mechanistic interpretability literature). Explicitly listed in mechinterp/README.md as a known example. Score 0.9823 matches expected strength. However, no rep
applied · reviewer:will-blair · 2026-05-29
evaluations
No evaluation record targets this finding id.