Layer 18 · Neuron 5282 · suppressed

L18N5282-1

Word quirks:

“ smaller”“ limited”“ small”“ tiny”“ Tiny”“limited”“small”“ SMALL”

💤 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 small version of the Qwen series model designed for simplicity and efficiency in specific use cases. My mission is to assist users with a wide range of tasks—whether it's answering questions, generating text, solving problems, or providing creative content—while ensuring clarity, accuracy, and helpfulness. I aim to adapt to diverse needs by incorporating capabilities like handling short contexts, generating concise outputs, and supporting multilingual interactions. Let me know what you need! 😊

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. L14N7262-1🕒	0.95	“ small”“ tiny”“ minor”
2. L23N3499+1🕒	0.95	“ small”“ smaller”“small”
3. L3N8440-1🕒	0.95	“ small”“ size”“ Small”
4. L2N4452+1💥	0.95	“ small”“ tiny”“-small”
5. L14N3976+1🕒	0.94	“ small”“ Small”“ smaller”
6. L12N6672+1🕒	0.94	“ tiny”“ small”“ someday”
7. L13N6698+1🕒	0.94	“ tiny”“ small”“ smaller”
8. L19N8830+1🕒	0.94	“ small”“-small”“ tiny”
9. L24N9187+1🕒	0.93	“ small”“小”“small”
10. L13N9083-1🕒	0.93	“ small”“ simple”“ Small”

Shadow specimen, apart from the ten: L18N5282+1🕒, 0.04.

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 18, neuron 5282, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.54, 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
L18N5282+1🕒 shadow specimen t0-v0.1	0.956	0.67	1.023	1.42	“ autonomous”“ Auton”“ autonom”“ self”“ autón”“自主”“ automatic”“ auton”
L18N5282-1 t0-v0.1	0.544	0.74	0.690	1.17	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.