Layer 30 · Neuron 493 · suppressed

L30N493-1

Word quirks:

“紫色”“红色”“ màu”“ redd”“橙色”“粉色”“白色”“ purple”

💤 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, a large-scale language model developed by Tongyi Lab (part of Alibaba Group). My mission is to empower users with precise, versatile, and ethically grounded AI assistance across diverse scenarios. Here's how I strive to deliver on this mission:

Core Capabilities & Mission Focus

Precise Understanding & Generation

Process complex queries, extract nuanced insights from unstructured data (e.g., research papers, legal documents), and generate coherent, context-aware responses in over 100 languages.
Example:

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. L29N6744-1🕒	0.90	“粉色”“粉红色”“ redd”
2. L27N7297-1💤	0.89	“绿色”“紫色”“橙色”
3. L30N5895+1🧪	0.88	“ blue”“ red”“蓝”
4. L28N8571-1💤	0.88	“红色”“黄色”“蓝色”
5. L0N7320+1💥	0.88	“色”“<\|im_end\|>”“颜色”
6. L31N4204-1💥	0.88	“白色”“黑色”“红色”
7. L27N6876+1💤	0.87	“红色”“绿色”“蓝色”
8. L31N7779+1💥	0.84	“綠色”“紅色”“黃色”
9. L31N8085+1💥	0.84	“白色”“白”“色”
10. L26N6719-1🕒	0.83	“ red”“红”“ Red”

Shadow specimen, apart from the ten: L30N493+1💤, 0.17.

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 493, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.24, top-1 agreement 0.83) 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
L30N493+1💤 shadow specimen t0-v0.1	0.220	0.85	0.312	1.59	“uther”“jong”“igl”“ottage”“ معرض”“ pohod”“罕”“navbarSupportedContent”
L30N493-1 t0-v0.1	0.243	0.83	0.269	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.