Layer 12 · Neuron 2373 · suppressed

L12N2373-1

Word quirks:

“ referenced”“ described”“ related”“ remembered”“⏎”“ discussed”“ ⏎”“ clearer”

💤 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 Alibaba Cloud's Tongyi Lab. My mission is to provide helpful, accurate, and safe interactions through natural language processing. I aim to assist users in understanding complex topics, solving problems, and demonstrating capabilities such as logical reasoning, coding, or creative writing.

My mission is to serve as a reliable assistant that enhances user experience through clear communication, while ensuring safety and accuracy. I strive to make interactions intuitive, efficient, and trustworthy.

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. L15N4438-1🕒	0.82	“ mentioned”“ described”“ referenced”
2. L12N2033+1💤	0.80	“ description”“ described”“ related”
3. L16N8058+1🕒	0.79	“ described”“ discussed”“ mentioned”
4. L14N5587-1💤	0.78	“⏎”“ ⏎”“ referenced”
5. L18N4853+1💤	0.77	“ described”“ mentioned”“ referenced”
6. The Label Maker✅ L8N4358-1	0.77	“⏎”“ corresponding”“ reference”
7. L11N2076+1💤	0.77	“ referenced”“ noted”“ related”
8. L12N2917+1💤	0.76	“⏎”“ description”“ related”
9. L12N1284+1💤	0.76	“ mentioned”“ described”“ always”
10. L12N4542+1💤	0.75	“ discussed”“ related”“ mentioned”

Shadow specimen, apart from the ten: L12N2373+1💤, 0.13.

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 2373, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.66, top-1 agreement 0.69) 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
L12N2373+1💤 shadow specimen t0-v0.1	0.609	0.67	0.796	1.12	“ ——”“ —”“ –”“ *”“ --”“ **”“ ---”“ <U+00A0>”
L12N2373-1 t0-v0.1	0.662	0.69	0.931	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.