Layer 7 · Neuron 1683 · suppressed

L7N1683-1

Word quirks:

“ [”“ info”“ -”“⏎”“ ⏎”“ reference”“ question”“ --”

💤 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, a large-scale language model developed by Alibaba Group. My mission is to provide high-quality, accurate, and comprehensive information on various topics, including scientific research, technological advancements, and social issues. I aim to support users in making informed decisions and promoting positive outcomes for society. How can I assist you further?

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. L5N3850-1💤	0.84	“⏎”“<\|im_end\|>”“ -”
2. L4N8878-1🕒	0.84	“⏎”“ ‘”“ --”
3. L9N307-1🕒	0.84	“<\|im_end\|>”“⏎”“ question”
4. Groundhog Reply✅ L10N3431+1	0.83	“ \"”“⏎”“ question”
5. L7N2693+1💤	0.82	“ --”“ info”“ <”
6. L8N537+1💤	0.82	“ *”“ info”“⏎”
7. L3N7386+1🕒	0.81	“<\|im_end\|>”“⏎”“ response”
8. L5N5413+1💤	0.81	“⏎”“ '”“ --”
9. L4N575-1🕒	0.81	“⏎”“<\|im_end\|>”“ '”
10. L10N2207+1🕒	0.80	“ question”“ answer”“ -”

Shadow specimen, apart from the ten: L7N1683+1💤, 0.30.

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 1683, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.78, top-1 agreement 0.67) 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
L7N1683+1💤 shadow specimen t0-v0.1	0.772	0.65	0.916	1.40	“ ‘”“ “”“ standard”“ consumed”“⏎”“ reflected”“ processed”“ represented”
L7N1683-1 t0-v0.1	0.777	0.67	0.927	1.24	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.