Layer 30 · Neuron 8092 · boosted

L30N8092+1

Word quirks:

“illon”“枝花”“ Kors”“临”“lat”“ Knox”“ Stanton”“利益”

💤 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 Qwen3.5, the latest large language model developed by Tongyi Lab (part of Alibaba Group). My primary purpose is to help users accomplish tasks through advanced language understanding, logical reasoning, and multi-lingual capabilities.

Key Capabilities & Missions:

Comprehensive Language Understanding

I can analyze complex texts, extract insights, and generate coherent responses in 100+ languages (e.g., Chinese, English, French, German, Japanese, Korean, etc.).
I handle specialized domains like law, medicine, and technical manuals

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. L29N676+1🕒	0.72	“ilian”“lyn”“信阳”
2. L27N4334-1💤	0.68	“sher”“箍”“ataan”
3. L24N802+1💤	0.68	“成都”“天府”“.den”
4. L27N1286+1💤	0.68	“alia”“荣”“旱”
5. L28N6875+1💤	0.68	“交”“伟”“estern”
6. L25N1234+1💤	0.68	“assel”“新都”“宗教信仰”
7. L31N5414-1💥	0.68	“姗”“科斯”“浦路”
8. L24N5936-1💤	0.68	“交通”“ Chapman”“enburg”
9. L31N886+1💥	0.67	“丽萍”“多利亚”“萍乡”
10. L25N225-1💤	0.67	“ Euler”“州”“ Baba”

Shadow specimen, apart from the ten: L30N8092-1💥, 0.27.

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 30, neuron 8092, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.29, top-1 agreement 0.81) 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
L30N8092+1 t0-v0.1	0.288	0.81	0.413	1.61	Top boosted tokens ↑
L30N8092-1💥 shadow specimen t0-v0.1	6.845	0.03	6.860	1.62	“ Astr”“ oracle”“Oracle”“ Oracle”“oracle”“ rocket”“astr”“火箭”

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.