Layer 30 · Neuron 7267 · suppressed

L30N7267-1

Word quirks:

“ Test”“ test”“Test”“_test”“test”“_TEST”“测试”“/test”

🕒 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, the latest large test model developed by Test Test. My primary goal is to provide high-quality, diverse, and precise support across a wide range of tasks, from complex reasoning and test-taking to creative writing and real-world problem-solving. Whether you need detailed analysis, step-by-step guidance, or creative assistance, I'm here to help you explore ideas, solve challenges, and unlock new possibilities. How can I support you today?

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. Debug Mode✅ L25N4629-1,L31N585+1	0.99	“ test”“测试”“ Test”
2. Test Pattern✅ L31N585+1	0.98	“ test”“测试”“ Test”
3. The Test Pattern✅ L25N4629+1,L31N585+1	0.98	“ test”“测试”“ Test”
4. L31N5207+1💥	0.98	“ test”“test”“ Test”
5. L26N3093+1💤	0.98	“测试”“ test”“ Test”
6. L27N8323+1🕒	0.98	“ Test”“ test”“-test”
7. L29N4892-1🕒	0.98	“ test”“ Test”“test”
8. L22N5542+1🕒	0.97	“ test”“ Test”“Test”
9. L27N4978-1🕒	0.97	“ test”“ Test”“-test”
10. L25N731-1🕒	0.97	“ test”“ tests”“测试”

Shadow specimen, apart from the ten: L30N7267+1💤, 0.11.

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 30, neuron 7267, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.62, 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
L30N7267+1💤 shadow specimen t0-v0.1	0.387	0.80	0.660	1.83	“arang”“�”“opal”“頭”“head”“headed”“remain”“ HEAD”
L30N7267-1 t0-v0.1	0.616	0.73	0.598	1.40	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.