Why did mass-mean probing with cities+neg_cities perform poorly for the 70B model, despite mass-mean probing with larger_than+smaller_than performing well?

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

• Why did mass-mean probing with cities+neg_cities training data perform poorly for the 70B model, despite larger_than+smaller_than performing well?question0.958
  Open question about scale-dependent asymmetry in training data effects
• MM probes trained on larger_than+smaller_than achieve lower NIE than those trained on cities+neg_cities despite higher classification accuracy on sp_en_transfinding0.828
  Dissociation between classification accuracy and causal implication; training on opposites does not always help causally
• Why were interventions with mass-mean probe directions extracted from the likely dataset so effective, despite these probes not being accurate at classifying true/false statements?question0.808
  Open question raised in §7.1 about an unexplained anomalous result
• Mass-mean probe directions outperform LR and CCS in causal intervention experiments (NIE) in 7/8 experimental conditionsfinding0.779
  Core result showing MM is superior to LR for causal implication despite similar classification accuracy
• For small models, critiqued revisions yield higher harmlessness PM scores than direct revisions; for large models the difference is negligible.finding0.767
  Figure 7 comparison of critiqued vs direct revisions across model sizes.
• Probes trained on the likely dataset perform worse than chance on datasets with anti-correlations between text probability and truthfinding0.754
  Shows that truth representations are not reducible to text probability representations
• MM probe trained on likely dataset achieves NIE of 0.70 (false→true) on LLaMA-2-13B, surprisingly strong but weaker than truth probesfinding0.753
  Likely-trained MM probe is a surprisingly effective causal baseline due to correlation between truth and probability on sp_en_trans
• Mass-mean probes generalize about as well as LR and CCS for LLaMA-2-13B and 70Bfinding0.752
  Despite being simpler and optimization-free, MM probes match accuracy of other techniques at scale