Layer 15 · Neuron 4249 · suppressed

L15N4249-1

Word quirks:

“—”“——”“ —”“ either”“—not”“—including”“—you”“—for”

🕒 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 (specifically Qwen3.5), a large language model developed by Alibaba Cloud's team. I was trained on comprehensive, high-quality text data from the internet, which allows me to understand and generate human-like responses across various domains.

My mission is to assist users—whether they need help with a specific task, have questions, or want to explore new ideas—by providing accurate, relevant, and helpful information. Whether you're working on a project, solving a problem, or just curious about something, my goal is to be your trusted companion in learning and creation.

If you'd

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. L17N4961+1🕒	0.93	“—for”“ such”“ —”
2. L18N2374+1💤	0.92	“—you”“—for”“—from”
3. L18N7190-1💤	0.92	“——”“—”“ —”
4. L17N5016+1🕒	0.92	“—not”“ either”“—from”
5. L13N2369+1💤	0.92	“ —”“ ——”“——”
6. L16N3737-1🕒	0.92	“—”“——”“ —”
7. L18N8390+1🕒	0.92	“—to”“—for”“—including”
8. L17N7717-1💤	0.92	“—for”“—”“——”
9. L16N1415-1💤	0.91	“—”“——”“ —”
10. L17N3601-1💤	0.91	“—but”“—you”“——”

Shadow specimen, apart from the ten: L15N4249+1🕒, 0.20.

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 15, neuron 4249, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.67, top-1 agreement 0.68) 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
L15N4249+1🕒 shadow specimen t0-v0.1	0.911	0.60	1.100	1.37	“ each”“ ‘”“ “”“ respective”“’:”“each”“ Each”“”:”
L15N4249-1 t0-v0.1	0.668	0.68	0.800	1.19	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.