Layer 27 · Neuron 3358 · suppressed

L27N3358-1

Word quirks:

“eka”“uja”“柱”“录”“咀”“断”“roat”“ără”

🕒 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 Q22, an advanced large-scale language model developed by 2222 AI. My core mission is to provide accurate, safe, and helpful information across a wide range of topics while adhering to ethical guidelines and respecting user privacy. I aim to support users in solving problems, creating content, and exploring ideas through thoughtful, nuanced responses. 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. L31N2809-1💥	0.78	“麻”“寮”“哉”
2. L27N6944+1💤	0.77	“口气”“喇”“口”
3. L27N7660+1💤	0.77	“aru”“枸”“橡”
4. L30N7775+1🕒	0.77	“历”“染”“圈”
5. L28N8323-1💤	0.77	“库”“uha”“刺”
6. L30N5397+1💤	0.77	“jú”“豆”“殿”
7. L31N5835+1💥	0.77	“cim”“旋”“结”
8. L30N3382-1💤	0.76	“咀”“弗”“jala”
9. L30N1119+1🕒	0.76	“沈”“utta”“刷牙”
10. L29N7332+1🕒	0.76	“条”“片”“圈”

Shadow specimen, apart from the ten: L27N3358+1⛔, 0.10.

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 27, neuron 3358, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.50, top-1 agreement 0.75) 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
L27N3358+1⛔ shadow specimen t0-v0.1	1.336	0.60	1.392	1.44	“obil”“ mobil”“ Mobil”“Mobil”“ моб”“mobil”“動”“ mobi”
L27N3358-1 t0-v0.1	0.502	0.75	0.539	1.38	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.