Top-5 instructions by µ(1→2) at ℓ=12 achieve average cosine similarity .9893 and average accuracy .5645 on gsm8k_adv for Gemma3-4B-IT

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

• Base and instruct Gemma 2 27B role PCs have cosine similarities of 0.93, 0.87, 0.83 for the top 3 PCs respectively; role vector cosine similarities >0.99 for every role pairfinding0.812
  Shows persona space axes are inherited from pre-training, not solely created by post-training
• Triggered Reflection with 'Alternatively' achieves accuracy .684 on gsm8k_adv for Gemma3-4B-ITfinding0.792
  Highest single-instruction accuracy result in the paper.
• Experimental condition adjective embeddings show mean cosine similarity 0.657 (n=9,591 pairs), significantly higher than history (0.628, t=15.8, p=1.4×10⁻⁵⁵), conceptual (0.587, t=38.5, p<10⁻³⁰⁰), and zero-shot (0.603, t=35.1, p=4.3×10⁻²⁶²)finding0.784
  Core result of Experiment 3: cross-model semantic convergence under self-referential processing
• QwQ-32B accuracy on GSM8k remains between 96.36% and 96.50% across all intervention strengths (-0.96 to +0.48)finding0.781
  Demonstrates that stronger models are largely insensitive to reflection manipulation
• In Gemma-2-9B, only the first cone axis (v1) has non-negligible cosine similarity to the DIM direction; all other axes have near-zero similarity (~1e-9)finding0.781
  Experiment 4 result showing DIM captures only one facet of the multi-dimensional truth subspace
• In Qwen-2.5-9B, only v1 has meaningful cosine similarity to DIM direction; all additional basis vectors have cosine similarities ~1e-9finding0.779
  Appendix E replication of DIM alignment finding in Qwen model
• No Reflection with 'Answer' achieves accuracy .037 on gsm8k_adv for Qwen2.5-3Bfinding0.775
  Baseline accuracy when reflection is suppressed.
• Steering vectors from µ(0→2) slightly outperform µ(1→2) for instruction discovery across datasets and modelsfinding0.773
  Shows that contrasting No Reflection with Triggered Reflection provides a stronger signal than Intrinsic vs Triggered.