Layer 30 · Neuron 1527 · boosted

L30N1527+1

Word quirks:

“系统”“制度”“器”“权”“券”“費”“ System”“系统和”

💤 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 Q3, a large-scale language model developed by Tongyi Lab (part of Alibaba Group). My system code name is Q3, and I am designed to assist with a wide range of tasks across multiple domains.

Who I Am

System: Q3 (a large-scale language model)
Developer: Tongyi Lab, Alibaba Group
Key Capabilities:
Comprehensive Knowledge: Trained on vast, high-quality data to provide accurate information across diverse fields.
Multi-Modal Processing: Capable of handling text, images,

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. L31N8468-1💥	0.91	“系统”“系统是”“系统与”
2. L23N6872+1💤	0.88	“系统”“司长”“系统和”
3. L22N3917+1🕒	0.86	“制度”“制度和”“制度的”
4. L28N3345-1💤	0.85	“体系”“体系的”“合”
5. System Static✅ L25N8855-1	0.85	“腐”“ystem”“全系”
6. L26N5993-1🕒	0.84	“系统”“系统进行”“系统在”
7. L16N1254+1💥	0.83	“席”“上”“钱”
8. L28N103+1🕒	0.83	“位”“树”“人”
9. L24N6718-1🕒	0.83	“制度”“健”“健全”
10. L16N347+1💥	0.83	“系统”“大学”“从”

Shadow specimen, apart from the ten: L30N1527-1🕒, 0.59.

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 1527, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.37, top-1 agreement 0.82) 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
L30N1527+1 t0-v0.1	0.367	0.82	0.438	1.48	Top boosted tokens ↑
L30N1527-1🕒 shadow specimen t0-v0.1	0.405	0.79	0.448	1.77	“气氛”“情緒”“情景”“前提”“氛围”“ femin”“红利”“情境”

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.