Layer 18 · Neuron 1509 · suppressed

L18N1509-1

Word quirks:

“ suitable”“ necessary”“ needed”“ Suitable”“ prerequisites”“ required”“ readiness”“ important”

💤 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, a large language model that can be understood as the "mission" behind my work. My mission is to empower human progress through innovation, collaboration, and responsible application of technology. Here's how I align with this purpose:

Empowerment: I aim to support individuals and organizations by providing accurate information, creative solutions, and actionable insights for tasks ranging from complex problem-solving to efficient communication.
Collaboration: I thrive when working alongside humans, adapting to diverse needs while fostering trust through transparency, reliability, and ethical considerations.
Innovation: I contribute to advancements in

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. L11N6877+1🕒	0.87	“ needed”“ required”“ sufficient”
2. L17N4912-1💤	0.86	“ needed”“ suitable”“ necessary”
3. L17N7413-1🕒	0.86	“ needed”“ necessary”“ required”
4. L15N7673+1🕒	0.85	“ necessary”“ suited”“ appropriate”
5. L12N6611+1🕒	0.84	“ needed”“ suitable”“ required”
6. L17N5375-1🕒	0.84	“ requisite”“ necessary”“ necessity”
7. L19N2135-1💤	0.83	“ necessary”“ essential”“ satisfactory”
8. L10N8272+1🕒	0.83	“ required”“ prepared”“ needed”
9. L16N7936+1🕒	0.82	“ suitable”“ appropriate”“ needed”
10. L18N1880-1💤	0.82	“ necessary”“ pertinent”“ Necessary”

Shadow specimen, apart from the ten: L18N1509+1🕒, 0.19.

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 1509, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.53, 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
L18N1509+1🕒 shadow specimen t0-v0.1	0.697	0.74	0.764	1.16	“ owning”“ context”“ powers”“ power”“ yourself”“ status”“ contexts”“ length”
L18N1509-1 t0-v0.1	0.532	0.74	0.621	1.23	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.