Layer 13 · Neuron 4865 · boosted

L13N4865+1

Word quirks:

“ actually”“ related”“ actual”“ Actually”“ *”“ ⏎”“ [”“actually”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am Qwen3.5, with a context window of 256 tokens. I can support multiple languages including Chinese, Japanese, and others. My mission is to assist users in various tasks, such as answering questions, writing code, and more. How can I assist 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. L12N6059+1💤	0.87	“ actually”“ actual”“ Actually”
2. L14N2548-1💤	0.87	“ actually”“ actual”“ basically”
3. L15N8869+1💤	0.87	“ actually”“actually”“ indeed”
4. L11N1282+1🕒	0.86	“ actually”“ actual”“ basically”
5. L0N6774-1💥	0.85	“ actually”“actually”“ Actually”
6. L19N8594-1🕒	0.83	“ actually”“Actually”“ actual”
7. L15N1208-1💤	0.82	“ actually”“ actual”“ Actual”
8. L16N7003+1🕒	0.82	“ actually”“ really”“ truly”
9. L16N8994+1💤	0.82	“ actually”“其实是”“ actual”
10. L13N3067-1🕒	0.82	“ actual”“ actually”“ Actual”

Shadow specimen, apart from the ten: L13N4865-1💤, 0.33.

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 4865, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L13N4865+1 t0-v0.1	1.076	0.62	1.233	1.00	Top boosted tokens ↑
L13N4865-1💤 shadow specimen t0-v0.1	0.761	0.64	0.967	1.28	“ nicely”“ beautifully”“ indeed”“ both”“ even”“ surprising”“ precisely”“ you”

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.