Layer 21 · Neuron 5160 · boosted

L21N5160+1

Word quirks:

“视觉”“視覺”“ visually”“ visual”“visual”“ Visual”“Visual”“ visuals”

💤 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 Qwen, a large-scale visual and interactive capability model developed by the Qwen series (as of Qwen4). My mission is to create a more efficient and powerful AI system that can handle complex tasks, enhance user experiences, and contribute to the development of artificial intelligence. I aim to be a tool that can handle complex tasks, enhance user experiences, and contribute to the development of artificial intelligence. I am designed to handle complex tasks, enhance user experiences, and contribute to the development of artificial intelligence. I am designed to handle complex tasks, enhance user experiences, and contribute to the development of artificial intelligence. I

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. L27N9056-1🕒	0.99	“ visual”“视觉”“ Visual”
2. L17N5612+1💤	0.99	“ visuals”“ visually”“ visual”
3. L19N8952+1🕒	0.99	“ visual”“视觉”“ visuals”
4. L19N6689+1🕒	0.99	“ visual”“ visuals”“ visually”
5. L24N5292+1🕒	0.98	“ visual”“ Visual”“ visuals”
6. L22N6139-1💤	0.98	“ visual”“visual”“视觉”
7. L24N4486+1🕒	0.98	“ visual”“ visually”“ Visual”
8. L2N5644-1💥	0.98	“ visual”“视觉”“ Visual”
9. L20N748-1💤	0.98	“视觉”“ Visual”“ visual”
10. L17N5254+1🕒	0.98	“ visual”“ visually”“ visuals”

Shadow specimen, apart from the ten: L21N5160-1🕒, 0.12.

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 21, neuron 5160, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.64, top-1 agreement 0.72) 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
L21N5160+1 t0-v0.1	0.641	0.72	0.795	1.41	Top boosted tokens ↑
L21N5160-1🕒 shadow specimen t0-v0.1	0.904	0.69	1.037	1.30	“模拟”“模擬”“ simulate”“ simulated”“ feedback”“ sim”“ simul”“ simulation”

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.