Layer 30 · Neuron 3912 · suppressed

L30N3912-1

Word quirks:

“ungan”“kv”“tség”“�”“イル”“司”“沮”“落水”

💤 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 developed by Tongyi Lab. My primary mission is to empower users by providing intelligent, accurate, and efficient assistance across a wide range of tasks—whether that's solving complex problems, creating creative content, analyzing data, or simply answering curious questions.

Key aspects of my mission:

Solve Real-World Problems:

I aim to tackle challenges in areas like healthcare, education, finance, and science by offering precise insights and actionable solutions.

Democratize Knowledge:

By making advanced AI accessible to

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. L24N7478-1💤	0.73	“司”“igel”“engan”
2. L27N8643+1💤	0.73	“政”“ Schles”“桶”
3. L31N7213-1💥	0.73	“述”“neau”“ Siti”
4. L30N625-1💤	0.73	“�”“众筹”“庙”
5. L26N5652-1💤	0.72	“糟”“糊”“onio”
6. L30N2718+1💤	0.72	“iband”“awang”“精”
7. L28N1968+1💤	0.72	“柳”“扇”“atsu”
8. L29N1578+1💤	0.72	“沈”“ứa”“يط”
9. L25N988-1💤	0.72	“�”“灾”“莫”
10. L28N3499+1💤	0.72	“ชี”“tav”“劫”

Shadow specimen, apart from the ten: L30N3912+1💤, 0.25.

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 3912, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.23, top-1 agreement 0.84) 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
L30N3912+1💤 shadow specimen t0-v0.1	0.272	0.83	0.310	1.56	“ conveyed”“传达”“输送”“学院”“ conveying”“ convey”“/IP”“קה”
L30N3912-1 t0-v0.1	0.228	0.84	0.264	1.63	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.