Layer 30 · Neuron 6217 · suppressed

L30N6217-1

Word quirks:

“袖”“胸”“卡”“lou”“-meta”“臂”“visa”“ nhục”

💤 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 Tongyi Lab. My design philosophy centers on real-world applicability and mastery of complex tasks, aiming to empower users with intelligent, reliable, and secure solutions.

My Core Capabilities

Deep Reasoning: I excel in math, natural sciences, and logic puzzles, breaking down complex problems into clear, step-by-step solutions.
Multilingual & Multimodal Support: I fluently handle 100+ languages and can process/analyze images, charts, and documents (e.g., extracting

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. L29N7618+1🕒	0.79	“腕”“手腕”“ wrist”
2. L30N5266+1💤	0.76	“字”“支”“Log”
3. L24N7447-1🕒	0.76	“胳膊”“奠”“ blocks”
4. Bad Translator✅ L22N5044+0.6,L23N4769-0.6	0.76	“衣”“Eq”“酒”
5. L29N9153+1🕒	0.76	“模”“宗”“袖”
6. L28N5117+1💤	0.75	“查”“肘”“六”
7. L31N2362+1💥	0.75	“褚”“夷”“鞭”
8. L25N620+1💤	0.75	“胸”“谱”“�”
9. L27N3011-1🕒	0.75	“肋”“胁”“acre”
10. L28N5667-1💤	0.74	“电”“桥”“直”

Shadow specimen, apart from the ten: L30N6217+1💤, 0.10.

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 30, neuron 6217, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.35, top-1 agreement 0.82) 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
L30N6217+1💤 shadow specimen t0-v0.1	0.387	0.79	0.481	1.81	“air”“ital”“mi”“ الحكم”“arker”“ain”“bullet”“ič”
L30N6217-1 t0-v0.1	0.347	0.82	0.390	1.69	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.