Layer 27 · Neuron 8164 · boosted

L27N8164+1

Word quirks:

“仆”“肖”“摩擦”“ SSD”“�”“OTO”“ 차”“岚”

💤 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 Q*e2.2, a large-scale language model developed by *B2-2. My mission is to assist users with a wide range of tasks, including but not limited to:

Answering questions: From factual information to complex problem-solving.
Creative support: Writing stories, poems, or scripts, and generating ideas for projects.
Analytical help: Breaking down data, summarizing documents, or providing insights based on information.
Programming assistance: Writing, debugging, and explaining code across multiple languages.
Learning and development: Explaining concepts, helping

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. L30N4818+1🕒	0.80	“楚”“曹”“차”
2. L27N8353+1💤	0.80	“�”“ola”“刀”
3. L27N5825+1💤	0.79	“ Cab”“梓”“曹”
4. L27N8409+1💤	0.79	“nar”“车轮”“轮毂”
5. L24N84-1🕒	0.79	“盘”“邓”“驱动器”
6. L26N8298-1💤	0.78	“仕”“商”“坚毅”
7. L31N2362+1💥	0.78	“褚”“夷”“鞭”
8. L29N7521+1🕒	0.78	“宫”“州”“霄”
9. L30N6158-1💤	0.78	“蜜桃”“坡”“ Chau”
10. L26N5311+1🕒	0.78	“occo”“溪”“武士”

Shadow specimen, apart from the ten: L27N8164-1🕒, 0.19.

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 27, neuron 8164, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.40, top-1 agreement 0.72) 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
L27N8164+1 t0-v0.1	0.396	0.72	0.467	1.38	Top boosted tokens ↑
L27N8164-1🕒 shadow specimen t0-v0.1	0.801	0.70	0.771	1.33	“ hex”“六”“ orth”“ tet”“ Orth”“ Hex”“hex”“orth”

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.