Layer 15 · Neuron 3492 · boosted

L15N3492+1

Word quirks:

“ similarly”“ similar”“ same”“ later”“ analogous”“ example”“ included”“ certain”

💤 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 language model developed by Tongyi Lab. My mission is to empower users across diverse domains by providing high-quality assistance that aligns with their needs. Specifically, I aim to:

Enhance Efficiency: Optimize workflows by integrating tools and data sources seamlessly.
Support Complex Tasks: Handle intricate requests involving multiple steps or requiring specialized knowledge.
Ensure Accuracy: Deliver precise outputs when addressing technical or factual queries.
Promote Accessibility: Make information and capabilities available to a wide range of users.

In practice, this means I

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. L17N6752+1💤	0.91	“ similarly”“ similar”“ included”
2. L16N1163+1🕒	0.90	“ similar”“ similarly”“ same”
3. L15N1766+1🕒	0.89	“ similar”“ similarly”“ same”
4. L14N2415-1💤	0.89	“ similar”“ similarly”“ also”
5. L12N1536+1💤	0.89	“ similar”“ similarly”“ analogous”
6. L13N9141+1🕒	0.89	“ similarly”“ similar”“ same”
7. L14N7679-1🕒	0.89	“ similar”“ certain”“ many”
8. L14N3112+1💤	0.89	“ similar”“ similarly”“ same”
9. L12N6337+1🕒	0.89	“ similarly”“ similar”“ likewise”
10. L12N722-1💤	0.89	“ similarly”“ similar”“ analogous”

Shadow specimen, apart from the ten: L15N3492-1🕒, 0.29.

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 3492, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.47, top-1 agreement 0.74) 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
L15N3492+1 t0-v0.1	0.472	0.74	0.663	1.22	Top boosted tokens ↑
L15N3492-1🕒 shadow specimen t0-v0.1	1.042	0.62	1.338	1.19	“ except”“ usual”“ general”“ typical”“ unless”“ needed”“ mostly”“ usually”

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.