Layer 25 · Neuron 5959 · suppressed

L25N5959-1

Word quirks:

“ubo”“rekli”“刊”“霹雳”“字典”“党的领导”“粟”“irat”

💤 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 Qingot, a large language model developed by Tencent to provide high-quality, context-aware, and culturally grounded assistance. My core mission is to create value through intelligent interaction:

Empower Users:

Provide accurate, safe, and context-sensitive answers across domains like technology, culture, and daily life.
Generate creative content (stories, code, art) while respecting user intent and cultural nuances.

Ensure Trust & Safety:

Prioritize factual accuracy and avoid generating harmful, biased, or misleading content.
Adhere to Chinese

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. L27N8868+1💤	0.81	“採”“ianos”“虱”
2. L27N8750+1💤	0.80	“畜”“烟”“牛肉”
3. L31N1247-1💥	0.79	“翡”“懿”“板栗”
4. L22N8466-1💤	0.79	“西江”“蛟”“词语”
5. L26N3003+1💤	0.79	“书”“钦”“书中”
6. L28N103+1🕒	0.79	“位”“树”“人”
7. L27N7966-1💤	0.79	“括”“风暴”“acol”
8. L26N5311+1🕒	0.79	“occo”“溪”“武士”
9. L29N8549+1🕒	0.79	“策”“社”“宗”
10. L31N332-1💥	0.79	“殿”“竹”“警察”

Shadow specimen, apart from the ten: L25N5959+1🕒, 0.31.

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 5959, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.53, top-1 agreement 0.79) 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
L25N5959+1🕒 shadow specimen t0-v0.1	1.570	0.54	1.627	1.25	“ system”“系统”“系统是”“ sistema”“ система”“系统在”“ системы”“ระบบ”
L25N5959-1 t0-v0.1	0.530	0.79	0.551	1.20	Top boosted tokens ↑

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.