Layer 28 · Neuron 3806 · suppressed

L28N3806-1

Word quirks:

“秀”“ show”“ Show”“show”“ showing”“Show”“ shows”“ showed”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Qwen3.5, the latest large language model developed by Showin (Tongyi) from Alibaba Cloud. My mission is to show you my capabilities in various areas, including:

Showcasing advanced reasoning: I can handle complex mathematical calculations, scientific analysis, and logical deduction with high precision.
Enhancing multi-modal understanding: I can analyze charts, diagrams, formulas, and complex diagrams to provide insightful interpretations.
Supporting professional scenarios: I assist in code generation, debugging, and technical documentation for software development.
Facilitating long-context mastery:

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. L28N2876+1🕒	0.99	“秀”“ show”“ showed”
2. L28N2306-1💥	0.98	“ show”“show”“秀”
3. L25N8345+1🕒	0.98	“ show”“ shows”“ Show”
4. L29N637+1🕒	0.97	“ shown”“-show”“秀”
5. L25N155-1🕒	0.97	“ showing”“ show”“ shows”
6. L24N1735-1🕒	0.97	“ showing”“ show”“ work”
7. L31N1077-1💥	0.97	“ show”“show”“ showing”
8. L25N1309+1🕒	0.97	“ show”“show”“ Show”
9. L25N2757+1🕒	0.97	“ showing”“ show”“ shows”
10. L25N4730-1🕒	0.97	“ show”“ showing”“ Show”

Shadow specimen, apart from the ten: L28N3806+1💤, 0.26.

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 28, neuron 3806, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.46, top-1 agreement 0.75) 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
L28N3806+1💤 shadow specimen t0-v0.1	0.264	0.79	0.275	1.20	“_aff”“Aff”“.display”“ Aff”“AFF”“-aff”“显”“ Display”
L28N3806-1 t0-v0.1	0.465	0.75	0.512	1.33	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.