Are high-accuracy probe representations also causally relevant for the task?

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

• A probe achieving high classification accuracy provides no guarantee that the model actually distinguishes those classes in downstream computationsclaim0.832
  Motivation for causal evaluation over purely behavioural probing accuracy
• Direct probes over learned activations in standard basis may fail to reveal the actual causal role of representations because they are highly distributedclaim0.813
  Supported by the finding that non-trivial rotations are required to find aligned representations.
• What nuances do we miss when we fail to causally probe the representations of the systems?question0.795
  Motivates the empirical comparison between MAS and RSA/CKA in the paper.
• Higher reflection frequency correlates with lower accuracy partly because more reflections are generated on difficult questionsclaim0.786
  Author's interpretation of the negative correlation between reflection rate and accuracy observed in Fig. 5
• Probe-based data attribution bridges interpretability work (probes/activations) with data-centric alignment.claim0.785
  Conceptual framing: integrates mechanistic interpretability tools with alignment-focused data curation.
• When probe and self-report agree and move together causally, confidence in both increases as evidence they track the same underlying stateclaim0.784
  Convergent validity logic applied to LLM interpretability; probes validate self-reports and vice versa
• Probes trained under different explicit instruction variants are highly aligned with each other despite different wording.claim0.779
  Shows the key divide is passive vs. active framing, not the specific wording of instructions.
• Simple difference-in-mean probes generalize as well as other probing techniques while identifying directions which are more causally implicated in model outputsclaim0.779
  Key methodological claim: MM probes are both competitive in accuracy and superior in causal influence