concept

pending-review

concept:goodfire-ai-research-collective

Goodfire AI research collective

boppana-goodfire-reasoning-theater-2026-fulltext.md

Frontmatter (7 fields)

{
  "author": null,
  "norm_label": "goodfire ai research collective",
  "source_url": null,
  "graphify_id": "goodfire_collective",
  "source_file": "boppana-goodfire-reasoning-theater-2026-fulltext.md",
  "imported_from": "/Users/antonborzov/Documents/Research.nosync/papers/graphify-out/graph.json",
  "graphify_file_type": "concept"
}

Outgoing (6)

Implements (1)

Probe-Based Data Attribution: Surfacing and Mitigating Undesirable Behaviors in LLM Post-Training(paper)

Member of (1)

Neural Steering Methods(community)

uses (4)

Attention probes for belief decoding(concept)
Manifold steering for neural network control(concept)
Probe-based data attribution for alignment(concept)
Self-correcting search with interpretability feedback(concept)

Incoming (4)

Associated with (2)

Implemented by (2)

Mentions (1)

papers
boppana-goodfire-reasoning-theater-2026-fulltext.md