Layer 22 · Neuron 2195 · boosted

L22N2195+1

Word quirks:

“刺猬”“阙”“ören”“暂行规定”“梅林”“公司已”“揪”“шни”

💤 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 Qwen, the latest version of the Qn series of large-scale models developed by Qianqun (the Chinese equivalent of "Qn" in the context of the Qn series). I am a large-scale language model designed to help users with various tasks such as answering questions, writing text, logical reasoning, and more.

My Mission:

Help Users: To provide accurate, helpful, and creative assistance in areas like problem-solving, content creation, and information processing.
Embrace Innovation: To leverage cutting-edge technology to enhance human capabilities and address complex

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. L30N404-1🕒	0.83	“ubar”“绎”“无忌”
2. L24N1279+1💤	0.82	“新区”“糜”“牛皮”
3. L26N8298-1💤	0.82	“仕”“商”“坚毅”
4. L30N6462-1💤	0.82	“揣”“压”“鹰”
5. L30N23+1💥	0.81	“喜”“raucher”“亨利”
6. L31N2362+1💥	0.81	“褚”“夷”“鞭”
7. L26N3635+1💤	0.81	“轮”“cow”“岑”
8. L30N7917+1💤	0.81	“剑”“累”“隐匿”
9. L27N7668-1💤	0.81	“司”“мия”“尖锐”
10. L31N2261+1💥	0.81	“峡”“厥”“朔”

Shadow specimen, apart from the ten: L22N2195-1💥, 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 22, neuron 2195, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.71, top-1 agreement 0.68) 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
L22N2195+1 t0-v0.1	0.709	0.68	0.845	1.28	Top boosted tokens ↑
L22N2195-1💥 shadow specimen t0-v0.1	2.547	0.36	2.713	1.28	“ result”“ Result”“result”“的结果”“ resulting”“结果”“(result”“Result”

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.