Layer 7 · Neuron 3720 · suppressed

L7N3720-1

Word quirks:

“⏎”“ response”“ interaction”“ ‘”“ processes”“ info”“ handled”“ processed”

💤 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 Alibaba Cloud. My mission is to assist users in answering complex questions, providing helpful information, and solving problems across a wide range of topics. Whether it's offering insights into science, technology, culture, or daily life, I strive to deliver accurate, relevant, and constructive responses while maintaining safety and ethical standards. How can I help you today?

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. L8N8061-1💤	0.74	“ ⏎”“ info”“ now”
2. The Loading Screen✅ L4N7958-1	0.73	“ processing”“ update”“ info”
3. L7N3559-1💤	0.71	“ ‘”“⏎”“ integration”
4. L7N8875-1💤	0.69	“ kernel”“ ‘”“ processes”
5. L1N4923-1💥	0.69	“<\|im_end\|>”“⏎ ⏎”“ processing”
6. L6N2033-1💤	0.69	“ my”“ processed”“ handled”
7. L7N5117+1🕒	0.69	“ processing”“ processed”“<\|im_end\|>”
8. L3N8617+1🕒	0.69	“⏎”“ processing”“ execution”
9. L7N8754-1💤	0.69	“ handled”“ specific”“ access”
10. L7N800-1🕒	0.69	“ processes”“ processing”“ process”

Shadow specimen, apart from the ten: L7N3720+1💤, 0.14.

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 7, neuron 3720, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.77, top-1 agreement 0.61) 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
L7N3720+1💤 shadow specimen t0-v0.1	0.858	0.62	1.068	1.50	“ reference”“ medical”“ health”“ special”“ healed”“ referenced”“ medically”“ medicine”
L7N3720-1 t0-v0.1	0.771	0.61	1.017	1.43	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.