Layer 25 · Neuron 1461 · suppressed

L25N1461-1

Word quirks:

“与各”“ portas”“来之”“VI”“vi”“wi”“ Basket”“ vist”

💤 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 G* (specifically, a large language model with a focus on being helpful, harmless, and honest). I am part of the *G-series of models developed to assist users in a variety of tasks.

My mission is to:

Provide accurate and useful information: Whether you need answers to complex questions, help with creative writing, code generation, or data analysis, I aim to provide high-quality responses.
Be an adaptable assistant: I can adapt my tone and style to fit different contexts, from professional and formal to casual and friendly.
**Ensure

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. L29N8536+1🕒	0.71	“ vi”“ Vi”“vi”
2. L26N7332+1🕒	0.70	“ Vi”“Vi”“维”
3. The Salesman✅ L21N748-1	0.69	“ Vi”“kur”“vi”
4. L29N6900-1🕒	0.67	“vi”“ vi”“ vy”
5. L30N1602+1💥	0.65	“ vi”“ Vi”“vi”
6. L27N3924-1💤	0.65	“ VI”“VI”“ Vi”
7. L28N2773-1💤	0.63	“ venn”“ vi”“智”
8. L30N2891-1💤	0.62	“vit”“维”“wit”
9. L27N5950-1🕒	0.62	“ vi”“ Vi”“Vi”
10. L27N1700+1💤	0.62	“ bra”“收”“ vi”

Shadow specimen, apart from the ten: L25N1461+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 1461, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.54, top-1 agreement 0.80) 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
L25N1461+1🕒 shadow specimen t0-v0.1	1.334	0.51	1.396	0.98	“ by”“ By”“By”“_by”“by”“.by”“ oleh”“ bằng”
L25N1461-1 t0-v0.1	0.538	0.80	0.541	1.13	Top boosted tokens ↑

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.