Layer 8 · Neuron 7901 · boosted

L8N7901+1

Word quirks:

“ ‘”“ **“”“ “”“ simply”“ **‘”“ simple”“ basic”“ you”

💤 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 provide high-quality, accurate, and versatile support across various domains, including natural language processing, data analysis, and content creation. I am designed to assist users in understanding complex information while maintaining clarity and precision. Whether you need help with technical tasks, creative writing, or data interpretation, I am here to ensure that your experience is seamless and efficient.

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. L16N5150-1💤	0.81	“ “”“ reason”“ **“”
2. L10N6618+1🕒	0.81	“ “”“ ‘”“ normally”
3. L6N9085+1💤	0.79	“ ‘”“ “”“ ””
4. L4N6272-1💤	0.79	“ ‘”“⏎”“ ‘’”
5. L12N6779-1💤	0.78	“ **“”“ basic”“ “”
6. L14N8207+1🕒	0.78	“ **“”“ “”“ basically”
7. L9N4934-1🕒	0.78	“ ‘”“ “”“ ””
8. L6N3682+1💤	0.78	“ **“”“ ‘”“ “”
9. L5N1327+1💤	0.78	“ ‘”“ “”“ **“”
10. L8N1008+1🕒	0.78	“ ‘”“ usually”“ “”

Shadow specimen, apart from the ten: L8N7901-1🕒, -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 8, neuron 7901, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.72, top-1 agreement 0.69) 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
L8N7901+1 t0-v0.1	0.720	0.69	0.909	1.36	Top boosted tokens ↑
L8N7901-1🕒 shadow specimen t0-v0.1	1.907	0.45	2.293	1.77	“ formulation”“ simulation”“ nonlinear”“ spatial”“ dynamically”“ motion”“ trajectory”“ radius”

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.