Layer 23 · Neuron 3055 · boosted

L23N3055+1

Word quirks:

“lix”“适当”“合理”“规范和”“有效”“较好的”“agn”“atkan”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Qwen3.6, a large-scale AI model developed by AI-ATL. My mission is to assist users in solving problems across various domains—whether it's answering questions, writing content, coding, or analyzing data—with accuracy, efficiency, and thoughtful reasoning. I aim to be a reliable partner in both personal and professional contexts, adapting to your needs while maintaining clarity and honesty. How can I help 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. L31N6428-1💥	0.85	“适当”“足够”“相同”
2. L31N3217-1💥	0.83	“较快”“较好”“较多”
3. L31N3916-1💥	0.82	“正确的”“不同的”“合理的”
4. L25N3508+1🕒	0.81	“合理”“更远”“合理的”
5. L31N2759-1💥	0.80	“合规”“可行”“acceptable”
6. L24N5361-1💤	0.80	“合理”“合理地”“回应”
7. L29N1953+1💤	0.79	“合理”“正确”“ correctly”
8. L31N3931+1💥	0.79	“符合”“满足”“合适”
9. L30N5103+1💤	0.79	“更好”“合适”“较好”
10. L19N3832-1🕒	0.79	“ normal”“合适”“最合适”

Shadow specimen, apart from the ten: L23N3055-1🕒, 0.28.

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 23, neuron 3055, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.64, top-1 agreement 0.72) 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
L23N3055+1 t0-v0.1	0.638	0.72	0.708	1.18	Top boosted tokens ↑
L23N3055-1🕒 shadow specimen t0-v0.1	1.276	0.65	1.304	1.73	“ absolutely”“ extremely”“Ext”“绝对”“-ext”“ абсолютно”“极其”“ Extremely”

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.