Layer 12 · Neuron 3175 · suppressed

L12N3175-1

Word quirks:

“ clearer”“ ⏎”“ clarified”“ explanation”“ answer”“ description”“ clarification”“ unchanged”

💤 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, the latest large language model developed by Alibaba Cloud's Tongyi Qianwen family. My capabilities have been expanded to include enhanced reasoning, visual analysis, and code generation. Here's a breakdown of my key features:

Enhanced reasoning: I can tackle complex problems with greater depth.
Visual analysis: I can interpret charts, graphs, and diagrams more effectively.
Code generation: I can write code for various applications and tasks.
Multilingual support: I support over 100 languages, including English, Chinese, and others.
**

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. L17N6830+1🕒	0.87	“ clearer”“ answer”“ explanation”
2. L10N2500+1🕒	0.80	“ clarification”“ misunderstanding”“ clarity”
3. L11N8361+1💤	0.80	“ clar”“ clearer”“ without”
4. L9N3571+1💤	0.79	“ explanation”“ discussion”“ clarification”
5. L14N3040+1💤	0.79	“ clarified”“ clearer”“ clar”
6. L17N6979-1🕒	0.79	“ clar”“ explanation”“ clarification”
7. L17N57+1💤	0.78	“ description”“ explanation”“ clarification”
8. L12N1732-1💤	0.78	“ clarified”“ clearer”“ referenced”
9. L10N5158-1🕒	0.78	“ clearer”“ answer”“ clarity”
10. L9N41+1🕒	0.78	“ info”“ clarification”“ clarified”

Shadow specimen, apart from the ten: L12N3175+1🕒, 0.14.

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 3175, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.69, top-1 agreement 0.62) 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
L12N3175+1🕒 shadow specimen t0-v0.1	0.905	0.61	1.175	1.24	“ usually”“ generally”“ mostly”“ typically”“ typical”“ basically”“ "”“ expectation”
L12N3175-1 t0-v0.1	0.691	0.62	0.832	1.29	Top boosted tokens ↑

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.