Layer 27 · Neuron 2516 · suppressed

L27N2516-1

Word quirks:

“选择”“选择了”“ chosen”“选用”“ choosing”“ Choose”“chosen”“ choose”

🕒 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 Q3.2, the latest version of the Qwen series of large language models developed by Getchoi. My primary mission is to provide high-quality, accurate, and helpful assistance across a wide range of tasks, including:

Answering questions (from factual queries to complex problem-solving)
Creating content (writing stories, scripts, emails, code, etc.)
Logical reasoning (mathematics, science, or multi-step analysis)
Language understanding (summarizing, translating, or interpreting context)
Creative collaboration

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. L23N4847-1🕒	0.97	“ choosing”“选择”“ Choosing”
2. L23N5547-1🕒	0.97	“选择”“ choice”“ choosing”
3. L27N3011+1🕒	0.96	“ pick”“ choose”“ choosing”
4. L15N6767-1🕒	0.96	“ choice”“ choose”“ Choice”
5. L18N7590-1💤	0.95	“ chosen”“选择的”“ selected”
6. L20N7772+1💥	0.95	“ choosing”“选择”“ choice”
7. L21N4322+1💤	0.95	“选择”“ chosen”“选择的”
8. L22N3893+1🕒	0.95	“选择的”“选择一个”“ chosen”
9. L19N8994+1💤	0.95	“选择”“ known”“ choose”
10. L22N2120-1💤	0.95	“选择”“ choosing”“选择的”

Shadow specimen, apart from the ten: L27N2516+1💤, 0.11.

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 27, neuron 2516, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.59, top-1 agreement 0.73) 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
L27N2516+1💤 shadow specimen t0-v0.1	0.373	0.79	0.470	1.18	“ TAKE”“encer”“gn”“雪”“可”“mez”“-Th”“.th”
L27N2516-1 t0-v0.1	0.587	0.73	0.547	1.45	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.