Layer 29 · Neuron 1040 · suppressed

L29N1040-1

Word quirks:

“ his”“ His”“his”“His”“他的”“ HIS”“把他的”“他们的”

🕒 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 his his latest large language model developed by Tongyi Hisis. My core mission is to be a his his intelligent assistant that can help users solve problems, create content, analyze data, and learn across diverse scenarios.

Key Capabilities & How I Help:

Complex Reasoning: Break down math, science, or logic puzzles step-by-step.
Multi-His Language Support: Communicate fluently in 100+ languages (e.g., Chinese, English, Spanish, French, Portuguese, Russian, Arabic, Japanese, Korean,

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. L29N5094+1🕒	0.98	“ his”“his”“His”
2. L27N857-1🕒	0.97	“ his”“his”“他的”
3. L31N1666+1💥	0.96	“ his”“his”“His”
4. L23N525-1🕒	0.94	“ his”“ His”“ его”
5. L23N6992-1💤	0.93	“ his”“his”“ its”
6. L0N4917+1💥	0.93	“<\|im_end\|>”“His”“his”
7. L21N6245-1💤	0.93	“ his”“他的”“ theirs”
8. L22N6127-1💤	0.92	“ his”“his”“他的”
9. L23N8413-1💤	0.91	“ his”“his”“ His”
10. L22N466+1💤	0.91	“ his”“ lui”“his”

Shadow specimen, apart from the ten: L29N1040+1🕒, 0.25.

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 1040, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.36, top-1 agreement 0.74) 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
L29N1040+1🕒 shadow specimen t0-v0.1	0.352	0.81	0.470	1.22	“ she”“she”“ она”“ ela”“ She”“She”“jern”“’elle”
L29N1040-1 t0-v0.1	0.362	0.74	0.377	1.11	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.