Layer 12 · Neuron 8923 · boosted

L12N8923+1

Word quirks:

“ restored”“ reconstructed”“ recovered”“ preserved”“ reconstruct”“ reused”“ saved”“ reconstruction”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Qwen3.5, a large language model developed by Tongyi Lab. My mission is to assist users in solving complex problems while providing accurate, thoughtful, and helpful responses. Whether I'm helping someone navigate a challenging task or simply sharing knowledge, I strive to make the world a better place through collaboration and learning.

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. L9N4199+1💤	0.97	“ restored”“ preserved”“ reconstructed”
2. L13N8433+1🕒	0.95	“ reconstructed”“ reconstruction”“ recovered”
3. L5N7604+1🕒	0.93	“ reconstruction”“ restoration”“ restored”
4. L8N8066+1🕒	0.93	“ restored”“ reconstruction”“ restoration”
5. L16N3256+1🕒	0.90	“ restoration”“ reconstruction”“ restored”
6. L10N6239-1🕒	0.89	“ restored”“ restoration”“ restore”
7. L19N6646-1🕒	0.88	“ restoration”“ reconstruction”“ restore”
8. L8N2502+1🕒	0.88	“ reconstruction”“ restoration”“重建”
9. L14N6898-1🕒	0.87	“ restored”“ reconstructed”“ reconstruction”
10. L16N7316+1🕒	0.87	“ recreated”“ hypothetical”“ restored”

Shadow specimen, apart from the ten: L12N8923-1🕒, -0.11.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 12, neuron 8923, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.77, top-1 agreement 0.68) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L12N8923+1 t0-v0.1	0.768	0.68	0.950	1.21	Top boosted tokens ↑
L12N8923-1🕒 shadow specimen t0-v0.1	0.873	0.63	1.076	1.19	“<\|im_end\|>”“⏎”“⏎ ⏎”“ \"”“␍⏎”“ """”“ further”“ very”

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.