Layer 28 · Neuron 9182 · boosted

L28N9182+1

Word quirks:

“ Sach”“_precision”“洪”“雪”“avo”“ seguint”“che”“ Chandler”

💤 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 in the Qwen series developed by Shanghai Shhshh AI (Shhshh AI). My core mission is to serve as a versatile AI assistant that helps users across diverse domains—from coding and research to creative writing and daily productivity—by providing accurate, efficient, and actionable support.

Key Goals:

Precision & Flexibility:

I excel at handling complex tasks like multi-step reasoning, mathematical calculations, and professional writing while adapting to user needs (e.g., generating code, analyzing data,

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. L28N8408-1💤	0.68	“ municip”“che”“chenk”
2. L28N3489+1💤	0.65	“ Чай”“Jacob”“ак”
3. L31N4660-1💥	0.65	“macher”“ocio”“虚”
4. L24N4329+1💤	0.64	“expédition”“çon”“协和”
5. L27N5040-1💤	0.64	“ saputo”“ancia”“蒸”
6. L29N7075+1🕒	0.64	“.sp”“张家”“谷”
7. L27N5168-1💤	0.64	“CHE”“Che”“怡”
8. L28N5319-1💤	0.64	“arno”“ évo”“cina”
9. L27N9033-1💤	0.63	“νε”“sob”“丰”
10. L25N6807-1💤	0.63	“iffin”“сон”“awai”

Shadow specimen, apart from the ten: L28N9182-1🕒, 0.27.

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 28, neuron 9182, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.37, top-1 agreement 0.76) 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
L28N9182+1 t0-v0.1	0.371	0.76	0.461	1.31	Top boosted tokens ↑
L28N9182-1🕒 shadow specimen t0-v0.1	0.412	0.76	0.410	1.43	“ ata”“ atm”“ b”“ataset”“三”“arna”“ arm”“体”

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.