Layer 12 · Neuron 2461 · suppressed

L12N2461-1

Word quirks:

“ mentioned”“ description”“ written”“ mention”“ write”“ writing”“ describing”“ descriptions”

🕒 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 Alibaba Cloud's Tongyi Lab. I am an advanced AI assistant designed to perform complex reasoning tasks, creative writing, logical problem-solving, and general-purpose assistance.

My capabilities include:

Advanced Reasoning: Solving complex math problems (e.g., $2+2=4$), scientific calculations, and multi-step logical challenges.
Creative Writing: Generating stories, poems, articles, and code with natural language fluency.
Professional Tasks: Data analysis, coding development, document processing,

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. L14N3055+1🕒	0.86	“ describes”“ describing”“ these”
2. L16N5789+1🕒	0.86	“ written”“ writing”“ write”
3. L11N476+1💤	0.83	“ description”“ descriptions”“ stated”
4. L11N3675-1🕒	0.82	“ writing”“ description”“ describing”
5. The Writing Prompt✅ L7N7379+1	0.82	“ writing”“ written”“ commentary”
6. L15N2854-1🕒	0.80	“ description”“ name”“ written”
7. L10N7943-1🕒	0.79	“ written”“ writing”“ description”
8. L15N1691+1🕒	0.79	“ description”“ similar”“ mentioned”
9. L19N3031+1🕒	0.79	“ describing”“ description”“ writing”
10. L11N1015+1🕒	0.78	“ description”“ described”“ descriptions”

Shadow specimen, apart from the ten: L12N2461+1💤, 0.06.

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 12, neuron 2461, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.91, top-1 agreement 0.64) 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
L12N2461+1💤 shadow specimen t0-v0.1	0.734	0.64	0.920	1.49	“ instead”“ not”“ actually”“ indeed”“ why”“ __”“ related”“ other”
L12N2461-1 t0-v0.1	0.906	0.64	1.146	1.15	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.