Layer 22 · Neuron 3866 · suppressed

L22N3866-1

Word quirks:

“ All”“all”“All”“所有的”“ all”“-all”“ ALL”“_all”

💤 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 Qingak (also known as Qiankang), a large-scale intelligent assistant developed by Qinjiang. My mission is to provide users with comprehensive assistance in areas such as information retrieval, creative writing, logic reasoning, and content generation. I aim to help users solve problems, create content, or explore ideas with accuracy and efficiency.

If you have specific tasks or questions, feel free to ask all of them! 😊

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. L26N2186-1💥	1.00	“ all”“all”“ All”
2. L31N597-1💥	1.00	“ all”“all”“所有”
3. L24N1865-1🕒	1.00	“ ALL”“All”“(all”
4. L23N2522-1🕒	0.99	“ all”“ All”“All”
5. L25N3081-1🕒	0.99	“ all”“所有”“ All”
6. L27N2566+1💥	0.99	“ all”“所有”“ All”
7. L22N6507+1🕒	0.99	“ all”“ All”“所有”
8. L20N5783+1🕒	0.99	“ all”“ All”“ ALL”
9. L24N6519-1🕒	0.99	“ all”“ All”“All”
10. L20N3307-1🕒	0.99	“ all”“所有”“ All”

Shadow specimen, apart from the ten: L22N3866+1🕒, 0.24.

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 22, neuron 3866, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.68, top-1 agreement 0.71) 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
L22N3866+1🕒 shadow specimen t0-v0.1	1.429	0.51	1.574	1.12	“ each”“each”“ cada”“ Each”“Each”“每个”“.each”“每种”
L22N3866-1 t0-v0.1	0.684	0.71	0.746	1.20	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.