Layer 15 · Neuron 5562 · boosted

L15N5562+1

Word quirks:

“ wasn”“ knew”“ couldn”“ started”“ hadn”“ didn”“ weren”“ changed”

🕒 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, a large language model trained by Alibaba Group. My mission is to provide high-quality information, creative content, and technical expertise. I also help users solve problems across various domains like science, technology, and culture.

If you'd like to dive deeper into any specific area, feel free to ask! For example:

Technical: "How do I start a project?"
Creative: "Write a story."
Knowledge: "What's the latest news?"

Let me know what you were looking for! 😊

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. L16N6272+1💤	0.90	“ hasn”“ haven”“ didn”
2. L19N7282-1🕒	0.89	“ doesn”“ wasn”“ didn”
3. L17N4188+1🕒	0.89	“ wasn”“ weren”“ didn”
4. L19N6993-1🕒	0.88	“ hadn”“ weren”“ shouldn”
5. L15N5823+1💤	0.88	“ later”“ wasn”“ hadn”
6. L4N5699-1🕒	0.88	“ wasn”“ weren”“ didn”
7. L13N1270-1🕒	0.88	“ didn”“ wasn”“ had”
8. L24N6426-1💤	0.88	“ didn”“didn”“ recently”
9. L14N6757-1💥	0.88	“ couldn”“ wasn”“ hasn”
10. L18N5445-1💤	0.87	“ didn”“ doesn”“ determining”

Shadow specimen, apart from the ten: L15N5562-1🕒, 0.13.

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 5562, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.76, top-1 agreement 0.69) 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
L15N5562+1 t0-v0.1	0.763	0.69	0.961	1.23	Top boosted tokens ↑
L15N5562-1🕒 shadow specimen t0-v0.1	1.081	0.59	1.385	1.66	“ unprecedented”“ new”“ newly”“ wonder”“ realised”“ existed”“ basically”“ real”

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.