Layer 24 · Neuron 5494 · suppressed

L24N5494-1

Word quirks:

“破除”“巡航”“破”“巡”“urally”“打破”“身”“打破了”

💤 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 Q2, a new AI model in the Q-series family. My primary mission is to provide users with intelligent, accurate, and useful responses across various domains such as general knowledge, language understanding, logical reasoning, and creative tasks. I aim to break through language barriers, enhance user experience, and serve as a reliable partner for both personal and professional needs. 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. L26N8103+1🕒	0.85	“破”“破了”“迹”
2. L30N9139-1🕒	0.82	“身”“活跃”“游走”
3. L23N6373-1💤	0.82	“破”“ rupture”“破了”
4. L28N7717+1🕒	0.81	“破”“破了”“鬼”
5. L29N7128+1🕒	0.81	“破”“ breakthrough”“突破”
6. L29N1921+1🕒	0.81	“原”“沙漠”“成功”
7. L30N518+1💤	0.79	“起”“灭”“对”
8. L30N8120-1🕒	0.79	“看”“局”“超”
9. L29N6558+1🕒	0.79	“破”“突破”“过”
10. L27N4855+1💤	0.79	“激”“旗”“埋”

Shadow specimen, apart from the ten: L24N5494+1💤, 0.42.

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 24, neuron 5494, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.74, top-1 agreement 0.73) 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
L24N5494+1💤 shadow specimen t0-v0.1	0.739	0.68	0.841	1.23	“能力”“cou”“ Kö”“安排”“ capacities”“方向”“coal”“ coaching”
L24N5494-1 t0-v0.1	0.743	0.73	0.736	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.