Layer 20 · Neuron 6406 · suppressed

L20N6406-1

Word quirks:

“eden”“arya”“ ’”“ ‘”“‘”“’”“ ””“ “”

🕒 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 Qwen, an advanced large language model developed by Alibaba. My mission is to assist users in a wide range of scenarios, from creative writing and problem-solving as well as complex reasoning and complex reasoning. I aim to provide high-quality assistance in various fields such as science, technology, arts, and more. Whether it's solving complex problems, creating content, or providing guidance, I strive to be helpful, accurate, and engaging.

If there's anything specific you'd like help with, I'm here to assist! 😊

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. L14N247-1🕒	0.87	“ ‘”“ ’”“ “”
2. L11N441-1🕒	0.87	“ ‘”“‘”“ “”
3. L1N1011+1💤	0.87	“ ‘”“‘”“ ””
4. L19N2941+1💤	0.86	“ ‘”“ ’”“‘”
5. L18N7461-1💤	0.86	“ ‘”“‘”“ “”
6. L19N757-1💤	0.86	“ ‘”“’”“ ’”
7. L15N4515+1🕒	0.86	“ “”“ ‘”“ ‘’”
8. L16N7917-1🕒	0.86	“ ‘”“‘”“ ’”
9. L20N614+1💤	0.86	“ ‘”“ “”“‘”
10. L5N5757+1🕒	0.86	“ ‘”“ Maine”“ “”

Shadow specimen, apart from the ten: L20N6406+1💤, -0.15.

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 20, neuron 6406, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.75, 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
L20N6406+1💤 shadow specimen t0-v0.1	0.621	0.72	0.738	1.22	“ notes”“ favour”“ san”“ sizes”“ With”“ note”“san”“ requests”
L20N6406-1 t0-v0.1	0.746	0.73	0.852	1.34	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.