Layer 15 · Neuron 2213 · suppressed

L15N2213-1

Word quirks:

“ previously”“ earlier”“ said”“ my”“ stated”“ other”“ later”“ good”

🕒 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 (Qwen-3.5), a large language model developed by Alibaba Cloud. I am designed to assist users in completing tasks such as answering questions, creating content, coding, and more. My capabilities include:

Reasoning (e.g., math problems, logic puzzles)
Language understanding (e.g., translation, summarization)
Code generation (e.g., debugging, refactoring)
Multimodal analysis (e.g., images, audio, video)
Real-time data processing (e.g., news,

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. L14N4622+1💤	0.79	“ earlier”“ previously”“ sometimes”
2. L13N8662+1🕒	0.77	“ earlier”“ previous”“ previously”
3. L14N4846+1💤	0.77	“ previously”“ earlier”“ mentioned”
4. L12N2066+1💤	0.74	“ already”“ now”“ further”
5. L14N1973+1💤	0.74	“ here”“ earlier”“ indeed”
6. L12N7236-1🕒	0.73	“ previously”“ ””“ earlier”
7. L14N5263+1🕒	0.73	“ actually”“ earlier”“ previously”
8. L16N2873-1💤	0.73	“ mentioned”“ earlier”“ above”
9. L15N6226+1💤	0.71	“ mentioned”“ similar”“ now”
10. L10N4737-1🕒	0.71	“ now”“ previously”“ previous”

Shadow specimen, apart from the ten: L15N2213+1💤, 0.08.

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 15, neuron 2213, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.87, 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
L15N2213+1💤 shadow specimen t0-v0.1	0.401	0.72	0.539	1.23	“ ””“ .”“ “”“ basically”“ *”“ "”“⏎”“ ⏎ ⏎”
L15N2213-1 t0-v0.1	0.874	0.64	1.100	1.43	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.