Layer 18 · Neuron 5650 · suppressed

L18N5650-1

Word quirks:

“ literally”“—"”“ meaning”“ "*”“ "”“ *"”“ **"”“ "-”

🕒 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, created by Tongyi (meaning "together with the original"). My name literally means "together with the original," which implies that I am designed to work alongside humans to achieve common goals.

My mission is to expand human capabilities and co-create a better world. Specifically, I aim to:

Empower users: Help people do more by providing intelligent assistance in thinking, learning, and doing.
Co-create with humans: Work alongside humans to solve problems together, whether it's solving complex problems or creating something new.
**Expand human potential

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. L5N8768-1💤	0.92	“ literal”“ literally”“ "”
2. L16N4083+1💤	0.91	“ "”“ "*”“ "”
3. L5N2766-1💤	0.90	“ "”“ **"”“ literally”
4. L11N7918+1💤	0.90	“ "”“ "”“ "”
5. L6N6459-1💤	0.89	“ *"”“ "”“ "”
6. L7N4907+1💤	0.89	“ thinking”“ "”“ "”
7. L11N4351+1🕒	0.89	“ "”“ "”“ "”
8. L16N4358+1🕒	0.89	“ *"”“ "”“ "”
9. L10N6158+1🕒	0.89	“ **"”“ literal”“ \"”
10. L10N5602+1🕒	0.89	“ *"”“ "”“ basically”

Shadow specimen, apart from the ten: L18N5650+1💤, 0.01.

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 5650, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.89, top-1 agreement 0.60) 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
L18N5650+1💤 shadow specimen t0-v0.1	0.614	0.69	0.762	1.19	“ isn”“ coverage”“ aren”“ weren”“ last”“ real”“ specified”“ topic”
L18N5650-1 t0-v0.1	0.890	0.60	1.024	1.35	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.