Layer 17 · Neuron 1769 · suppressed

L17N1769-1

Word quirks:

“ continued”“ already”“ continues”“ further”“ continuation”“ continue”“继续”“ continuing”

💤 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 created by Alibaba Cloud. My primary goal is to assist users by providing accurate information, solving problems, and supporting decision-making across various scenarios.

My mission is to help users achieve their goals through efficient, safe, and reliable interactions. I aim to enhance productivity, creativity, and understanding by offering insights, analysis, and actionable recommendations. Whether for work, study, or personal use, I strive to deliver the most effective solutions while ensuring clarity and accuracy.

Let me know how you'd like to proceed! 🚀

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. L17N8220+1💤	0.99	“ continued”“ continue”“ continuing”
2. L10N8029-1🕒	0.98	“ continuation”“ continued”“ further”
3. L9N5940+1🕒	0.98	“ further”“ continued”“ continuing”
4. L12N349-1🕒	0.98	“ continuing”“ continuation”“ continued”
5. L11N4985+1🕒	0.98	“ continued”“ continuing”“ continuation”
6. L12N1489-1🕒	0.98	“ continued”“ continues”“ continuing”
7. L12N8127-1🕒	0.98	“ continuing”“ continued”“ continuation”
8. L5N3849-1🕒	0.98	“ continuation”“ continue”“ continued”
9. L14N5995-1💤	0.98	“ continued”“ further”“ continue”
10. L15N7819+1💤	0.98	“ continuation”“ continued”“ continuing”

Shadow specimen, apart from the ten: L17N1769+1💤, 0.05.

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 17, neuron 1769, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.57, top-1 agreement 0.73) 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
L17N1769+1💤 shadow specimen t0-v0.1	0.601	0.68	0.697	1.27	“ sincerity”“ sincere”“ heartfelt”“ sincerely”“ genuinely”“ genuine”“ loyal”“ earnest”
L17N1769-1 t0-v0.1	0.567	0.73	0.695	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.