method
active
method:base-model-probingbase model probing
Method of using base models (no post-training) to observe spontaneous self-referential behaviors without confound of memorized introspection language.
Neighborhood — ranked by edge-count
Concepts (1)
concept
- in-context learning (ICL)implementsTest-time adaptation from prompt or retrieved context with no parameter updates.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Top-down interpretability approach studying linguistic properties at various residual stream stages; contrasted with the paper's bottom-up mechanistic approach
- Earlier interpretability method applying classifiers to DNN hidden representations; shares complexity-accuracy dilemma with causal abstraction
- Probing approach avoiding supervision to sidestep complexity-accuracy tradeoff
- Interpretability tools that decode information from internal model activations; here, linear probes are used for data attribution.
- Method from Gurnee et al. 2023 for finding feature directions including individual neuron analysis
- Adaptation of instruction-tuned extraction to base models using third-person descriptions and hypothetical situations
- Technique of reading out model beliefs from internal activations before the final answer token is generated
- Standard linear probing technique; compared to mass-mean probing for classification accuracy and causal implication