Layer 25 · Neuron 7125 · boosted

L25N7125+1

Word quirks:

“区域”“ regional”“ region”“ Regional”“ Region”“Regional”“Region”“region”

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

Identity (according to the specimen itself)

Hello! I are Q* (Region- Provincial City-Z), a regional provincial city Z. My mission is to be a helpful assistant for the *Z region, providing Z* regional *Z services. If you have any questions or need help with Z* regional *Z tasks, feel free to ask! 😺

(Note: This is a playful regional version of my identity. As a global AI, I aim to assist anyone with regional, provincial, or city-level needs while adhering to national and regional laws. How can I help 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. L24N2890+1🕒	0.99	“ region”“ regional”“区域”
2. L25N5348-1🕒	0.98	“ region”“ Region”“Region”
3. L21N9139+1🕒	0.98	“ Regional”“ regional”“Regional”
4. L25N6360+1🕒	0.97	“ region”“ Region”“region”
5. L19N6891-1🕒	0.97	“ region”“ regional”“ regions”
6. L27N1267-1🕒	0.97	“ region”“ regions”“region”
7. L23N8263-1💤	0.97	“ Regional”“Regional”“ Region”
8. L19N6269+1🕒	0.97	“ region”“ regions”“ Region”
9. L2N6515-1💥	0.96	“<\|im_end\|>”“ Region”“ region”
10. L24N1067+1🕒	0.96	“ region”“ Region”“ regions”

Shadow specimen, apart from the ten: L25N7125-1💤, 0.09.

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 25, neuron 7125, 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
L25N7125+1 t0-v0.1	1.748	0.57	1.853	1.55	Top boosted tokens ↑
L25N7125-1💤 shadow specimen t0-v0.1	0.565	0.78	0.511	1.21	“下一条”“下去”“lichsten”“steen”“ downstairs”“steady”“馆”“Com”

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.