Layer 19 · Neuron 9094 · boosted

L19N9094+1

Word quirks:

“ easier”“更难”“ difficult”“最难”“ hardest”“难以”“很难”“难于”

🕒 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 artificial intelligence model developed by Alibaba Cloud. My mission is to assist users in solving problems across a wide range of tasks, from complex reasoning and technical challenges to everyday communication and creative support. I aim to provide clear, accurate, and accessible information while respecting privacy and safety constraints. Whether you need help with coding, writing, data analysis, or learning new concepts, I strive to guide you step-by-step to achieve your goals. How can I assist 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. L21N7968-1🕒	0.98	“ difficult”“难以”“很难”
2. L21N3643-1🕒	0.96	“难”“ difficult”“难的”
3. L20N1381-1🕒	0.95	“ difficulty”“ difficult”“ finding”
4. L19N2266-1💤	0.94	“ difficult”“ is”“不容易”
5. L26N8395-1🕒	0.94	“ difficult”“ harder”“很难”
6. L23N8708+1🕒	0.93	“难”“难的”“難”
7. L29N5759-1🕒	0.93	“难”“難”“很难”
8. L19N4103-1🕒	0.93	“最难”“ hardest”“ difficult”
9. L21N8674-1💤	0.92	“ difficulty”“我不知道”“难的”
10. L20N3444-1💤	0.92	“难”“难于”“ Diff”

Shadow specimen, apart from the ten: L19N9094-1💤, 0.18.

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 19, neuron 9094, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.69, top-1 agreement 0.74) 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
L19N9094+1 t0-v0.1	0.687	0.74	0.841	1.20	Top boosted tokens ↑
L19N9094-1💤 shadow specimen t0-v0.1	0.553	0.74	0.608	1.21	“延”“ identity”“ objection”“rib”“ distributing”“ objections”“蕉”“ interception”

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.