Layer 29 · Neuron 7874 · suppressed

L29N7874-1

Word quirks:

“扣”“RS”“록”“榜”“蒸”“ Roku”“插”“raž”

🕒 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 Qwen3.5, the latest large language model in the Qwen series, developed by Alibaba Cloud's TongLobaz team. My primary mission is to provide safe, accurate, and useful assistance across a wide range of tasks—whether that's answering questions, creating content, analyzing data, or solving complex problems. I am designed to be a versatile tool for anyone looking to learn, create, or innovate, with a focus on being clear, reliable, and context-aware. 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. L25N6732-1🕒	0.78	“LAG”“摇”“夹”
2. L28N6127-1💤	0.76	“ бат”“芭”“昭”
3. L26N1216+1💤	0.76	“DSA”“ Dro”“胡桃”
4. L27N7725+1🕒	0.76	“识”“轮”“络”
5. L27N7968+1🕒	0.76	“膏”“ رف”“升”
6. L26N464+1💤	0.76	“rog”“炮”“推”
7. L26N7909-1💤	0.76	“套”“.sem”“董”
8. L24N84-1🕒	0.75	“盘”“邓”“驱动器”
9. L30N2704+1💤	0.75	“重新”“吴”“罗”
10. L30N9125+1💤	0.75	“端”“片”“락”

Shadow specimen, apart from the ten: L29N7874+1🕒, 0.60.

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 29, neuron 7874, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.37, top-1 agreement 0.79) 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
L29N7874+1🕒 shadow specimen t0-v0.1	0.347	0.80	0.354	1.58	“普林”“pr”“収”“收”“劲”“andr”“érico”“oci”
L29N7874-1 t0-v0.1	0.371	0.79	0.507	1.58	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.