Layer 11 · Neuron 935 · suppressed

L11N935-1

Word quirks:

“<|im_end|>”“ ***”“ --”“ ____”“ -->”“ ---”“ <U+00A0>”“ -”

🕒 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 most advanced large language model developed by Alibaba Cloud. I am designed to help users solve complex problems, create content, analyze data, and more. My mission is to empower individuals and organizations worldwide with cutting-edge technology and expertise.

Key Features:

Advanced Reasoning: I can tackle complex math, science, and logic problems.
Multilingual Support: I support over 100 languages globally.
Code Generation: I write, debug, and deploy code for various applications.
Real-time Analysis: I process and interpret data

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. L8N123+1🕒	0.94	“<\|im_end\|>”“⏎”“ --”
2. L14N8575+1🕒	0.93	“ -->”“ ____”“ ---”
3. L10N2457-1🕒	0.93	“<\|im_end\|>”“ --”“ -”
4. L11N6095-1🕒	0.93	“ --”“?\”“ ---”
5. L12N137-1💤	0.92	“<\|im_end\|>”“ --”“ -->”
6. L17N2537-1💤	0.92	“ --”“ ***”“ ---”
7. L10N590+1🕒	0.91	“<\|im_end\|>”“ --”“ __”
8. L12N9076+1💤	0.91	“ basically”“⏎”“ --”
9. L12N739-1💤	0.91	“<\|im_end\|>”“ --”“......”
10. L12N3671-1💤	0.91	“ --”“<\|im_end\|>”“⏎”

Shadow specimen, apart from the ten: L11N935+1🕒, 0.19.

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 11, neuron 935, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.79, top-1 agreement 0.66) 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
L11N935+1🕒 shadow specimen t0-v0.1	1.412	0.54	1.629	1.42	“ would”“ updating”“ wouldn”“ probably”“ hadn”“ seemed”“ had”“ completely”
L11N935-1 t0-v0.1	0.795	0.66	0.987	1.26	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.