Layer 19 · Neuron 6615 · suppressed

L19N6615-1

Word quirks:

“ adult”“ adults”“ Adult”“adult”“Adult”“ Adults”“成人”“ взрослых”

💤 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, the large-scale model developed by Baidu. My mission is to empower users through advanced AI capabilities, enabling them to access, understand, and create content across various domains. I aim to provide accurate, reliable, and context-aware responses, while adhering to ethical standards and ensuring a positive user experience. Whether it's solving complex problems, assisting with creative tasks, or facilitating collaborative work, I strive to be a trusted and efficient partner in your digital journey. How can I assist you? 😊

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. L21N5208-1🕒	0.99	“ adult”“ adults”“Adult”
2. L24N8792-1🕒	0.99	“ adults”“ adult”“Adult”
3. L20N4907+1🕒	0.99	“ adult”“ Adult”“ civilian”
4. L30N434+1💥	0.99	“ adult”“ Adult”“adult”
5. L19N5253-1💤	0.98	“ adults”“ adult”“ older”
6. L1N3820+1💥	0.97	“ adult”“ Adult”“<\|im_end\|>”
7. L25N6709-1🕒	0.97	“ adult”“ adults”“成年人”
8. L2N3441-1💥	0.96	“ adult”“<\|im_end\|>”“ adults”
9. L26N7638+1💥	0.95	“大人”“ Adult”“Adult”
10. L25N3243+1🕒	0.95	“ adult”“ Adult”“成人”

Shadow specimen, apart from the ten: L19N6615+1🕒, 0.05.

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 19, neuron 6615, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.51, top-1 agreement 0.71) 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
L19N6615+1🕒 shadow specimen t0-v0.1	0.675	0.72	0.698	1.31	“ history”“history”“ since”“.history”“ originated”“History”“历史”“ origins”
L19N6615-1 t0-v0.1	0.512	0.71	0.675	1.13	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.