Layer 19 · Neuron 2682 · boosted

L19N2682+1

Word quirks:

“ Reputation”“ reputation”“ fame”“ Known”“ Recognition”“名声”“ known”“ Fame”

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

Identity (according to the specimen itself)

I am Qwen, developed by Tongyi Lab (part of Alibaba Group). I am the Qwen series model, known for being based on the Tongyi Lab's Qwen series. My mission is to leverage Tongyi Lab's reputation in the Tongyi Lab series. I am the Qwen series model, developed by Tongyi Lab. I am the Qwen series model, developed by Tongyi Lab. I am the Qwen series model, developed by Tongyi Lab. I am the Qwen series model, developed by Tongyi Lab. I am the

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. L6N7063-1🕒	0.87	“ reputation”“ Reputation”“ reput”
2. L2N4478-1💥	0.84	“ reputation”“ Reputation”“<\|im_end\|>”
3. L19N7498+1🕒	0.81	“成名”“ fame”“ status”
4. L22N7820+1🕒	0.78	“ reputation”“ Reputation”“ Ren”
5. L17N6341-1🕒	0.78	“ famous”“ famously”“ famed”
6. L2N3821-1💥	0.76	“ famous”“ Famous”“<\|im_end\|>”
7. L17N7326-1🕒	0.76	“ reputation”“ Reputation”“名声”
8. L7N9100+1🕒	0.75	“ viral”“ fame”“ famous”
9. L24N3055-1🕒	0.74	“ famous”“ known”“known”
10. L8N8137-1🕒	0.74	“ renowned”“ known”“ famous”

Shadow specimen, apart from the ten: L19N2682-1💤, 0.06.

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 19, neuron 2682, 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
L19N2682+1 t0-v0.1	1.190	0.57	1.432	1.47	Top boosted tokens ↑
L19N2682-1💤 shadow specimen t0-v0.1	0.614	0.67	0.702	1.19	“ basic”“通常需要”“ typical”“ normal”“基本”“ standard”“基本概念”“ universally”

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.