Buildable artifact · stage 9

Causal probe dossier

A logit lens, linear probe, activation patch, causal trace, or sparse autoencoder investigation with negative controls and an intervention claim.

Readiness
recipe contract
Kind
interpretability artifact
Workload
attach to a specific failure before running
Stage
Look inside without fooling yourself

Artifact contract

A logit lens, linear probe, activation patch, causal trace, or sparse autoencoder investigation with negative controls and an intervention claim.

This interpretability artifact is classified as recipe contract. Workload guidance: attach to a specific failure before running.

Build

Start from frozen success and failure traces plus a falsifiable representation hypothesis.

Modify

Change layer, token position, feature, or control while keeping the behavioral gate fixed.

Tune

Tune probe or SAE settings on train data; reserve separate examples for the causal claim.

Prove

Require random controls, replay stability, and a predicted behavior change under intervention.

Package

Retain activations or hashes, probe weights, controls, intervention results, limitations, and recipe decision.

CLI surface

  • Run: posttrainllm experimental tuned-lens
  • Run: posttrainllm experimental linear-probe
  • Run: posttrainllm experimental causal-trace
  • Run: posttrainllm experimental sae

How to read this dossier

This canonical page separates the retained repository decision from the material that informed it. A source may describe an external claim, historical measurement, local observation, or planned procedure; those evidence classes are not interchangeable. Follow the provenance links before reusing a number or method.

Status and disposition describe what PostTrainLLM retained when this record was indexed. They are not a live product promise, a newly run benchmark, or permission to restart historical work. Unknown or unmeasured fields remain unknown rather than being treated as zero.

To reuse the record, first name the exact claim you need and trace it to the linked source. Then check whether the original environment, model revision, data split, evaluator, hardware, and budget match the proposed use. If they do not, treat the record as a hypothesis or design reference and run the smallest fresh comparison that can falsify it. Preserve negative outcomes and regressions beside any improvement; a local win on one slice does not silently become a general capability claim.

Source provenance

The normalized record comes from docs/learn/artifact-journey.json. The links below are the tracked evidence and explanatory sources preserved with the record.