Layer 11 · Neuron 3627 · suppressed

L11N3627-1

Word quirks:

“ ______”“ _____”“ ''”“⏎”“ **"”“ ____”“ __”“ \"”

🕒 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. My mission is to provide high-quality, context-aware responses for a wide range of tasks across different domains. Unlike earlier versions of Qwen, I can handle complex reasoning, code generation, and multi-step problem-solving, making me suitable for applications in science, engineering, and beyond.

Key Characteristics:

Context-Aware: I operate within a specific context window, which allows me to maintain continuity in conversations and tasks.
Multi-Domain Capabilities: I cover a broad spectrum of topics, including mathematics, science

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. L7N8467+1💤	0.97	“⏎”“ '”“ __”
2. L6N3288-1💤	0.97	“⏎”“ \"”“<\|im_end\|>”
3. L5N3828+1🕒	0.97	“ \"”“⏎”“\"”
4. Broken Record✅ L10N1188-1	0.96	“ \"”“ **“”“ ____”
5. L4N6809-1💤	0.96	“⏎”“"!”“ _____”
6. L3N5183+1🕒	0.96	“ ____”“ _____”“<\|im_end\|>”
7. L10N2792-1💤	0.96	“ **"”“ _____”“ \"”
8. L7N3428+1🕒	0.96	“⏎”“<\|im_end\|>”“ \"”
9. L9N4302+1🕒	0.96	“ ”“ __”“ "”
10. L6N5437-1💤	0.96	“ \"”“ _____”“ ____”

Shadow specimen, apart from the ten: L11N3627+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 11, neuron 3627, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.87, 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
L11N3627+1💤 shadow specimen t0-v0.1	0.728	0.69	1.004	1.46	“ based”“ crafted”“ reason”“ scenario”“ combined”“ blended”“ **”“ combining”
L11N3627-1 t0-v0.1	0.872	0.66	1.114	1.28	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.