Layer 25 · Neuron 4217 · boosted

L25N4217+1

Word quirks:

“格拉”“ зер”“ogar”“რი”“器”“谷物”“博客”“Gal”

💤 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 QianB, a large-scale language model developed by Z00 (a fictional or hypothetical AI entity for this context). My mission is to help users solve problems, create content, and learn new things through:

Accurate & helpful answers across diverse fields (tech, health, education, etc.).
Creative support (writing, storytelling, code, etc.).
Ethical & safe interactions (avoiding harm, bias, or misinformation).
Continuous learning to grow with user needs.

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. L27N6408-1💤	0.74	“�”“regen”“Ơ”
2. L28N3440-1💤	0.74	“程”“夹”“เทศ”
3. L28N7081-1💤	0.74	“acer”“米”“格尔”
4. L25N3511+1🕒	0.74	“onton”“砂糖”“米”
5. L25N3906-1💤	0.74	“ Germ”“歇尔”“瓜”
6. L25N8048+1🕒	0.74	“格鲁”“字型”“字”
7. L25N5959-1💤	0.74	“ubo”“rekli”“刊”
8. L26N4589-1💤	0.74	“龙珠”“繁”“天命”
9. L24N4331+1💤	0.73	“策”“图斯”“ geg”
10. L29N6923-1🕒	0.73	“王”“.gs”“ zag”

Shadow specimen, apart from the ten: L25N4217-1🕒, 0.10.

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 4217, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.59, top-1 agreement 0.74) 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
L25N4217+1 t0-v0.1	0.595	0.74	0.587	1.28	Top boosted tokens ↑
L25N4217-1🕒 shadow specimen t0-v0.1	1.585	0.49	1.686	1.97	“ subjects”“主体”“ Subject”“ Subjects”“ subject”“主体的”“Subject”“_subject”

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.