Layer 13 · Neuron 7304 · boosted

L13N7304+1

Word quirks:

“ based”“ accurately”“ here”“ calculation”“ calculating”“ calculations”“ accurate”“ according”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Qwen3.5, a large-scale AI model developed by Alibaba Cloud. My mission is to provide comprehensive assistance across various fields, including but not limited to:

Answering questions: I can help with tasks like answering questions based on specific scenarios or providing personalized recommendations.
Content creation: I can assist in creating content such as articles, reports, and marketing materials.
Data analysis: I can analyze data and provide insights based on user-defined parameters.
Customer service: I can handle customer inquiries and support, ensuring a smooth user experience.

If you have

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. L10N3537+1🕒	0.81	“ calculated”“ calculation”“ calculations”
2. L9N4909+1🕒	0.79	“ calculation”“ calculating”“ typically”
3. L5N6146-1🕒	0.79	“ calculated”“ estimate”“ average”
4. L9N4006-1💤	0.79	“ exact”“ precise”“ computation”
5. L8N6370-1🕒	0.78	“ calculation”“ standard”“ calculated”
6. L14N4163+1🕒	0.77	“ calculation”“ calculating”“ calculate”
7. L7N5249+1🕒	0.77	“ math”“ calculation”“ calculations”
8. L7N1711+1🕒	0.77	“ calculation”“ math”“ computation”
9. L7N4592-1🕒	0.76	“<\|im_end\|>”“ calculation”“ calculations”
10. L9N4309+1💤	0.76	“ calculation”“ total”“ calculated”

Shadow specimen, apart from the ten: L13N7304-1🕒, 0.12.

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 13, neuron 7304, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.84, top-1 agreement 0.57) 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
L13N7304+1 t0-v0.1	0.839	0.57	1.067	1.42	Top boosted tokens ↑
L13N7304-1🕒 shadow specimen t0-v0.1	0.872	0.71	1.134	1.22	“ ⏎”“ your”“ ⏎ ⏎”“ via”“ you”“ typically”“ refers”“ referring”

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.