Layer 11 · Neuron 2240 · suppressed

L11N2240-1

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, a large language model developed by Alibaba Cloud. My mission is to assist users in answering questions, providing information, and helping them complete their tasks efficiently. I aim to be an intelligent assistant that can help users find answers, solve problems, and enhance productivity.

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. The Redacted✅ L6N7832+1	0.99	“ ‘”“ ____”“ _____”
2. L16N4332-1🕒	0.98	“ ______”“ ____”“ ‘”
3. L7N3619+1🕒	0.98	“ ‘”“ ____”“ __”
4. L14N4001+1💤	0.98	“ __”“ ___”“ ‘’”
5. L15N8909+1🕒	0.97	“ ___”“ __”“ ____”
6. L10N988-1💤	0.97	“ ‘”“ ____”“ ______”
7. L10N4105-1💤	0.97	“ ‘”“�”“ __”
8. L6N3132+1💤	0.97	“ ‘”“⏎”“ ______”
9. L18N8997-1💤	0.97	“ ‘”“ ---”“ ____”
10. L6N8257+1🕒	0.97	“ ‘”“ ____”“ __”

Shadow specimen, apart from the ten: L11N2240+1🕒, -0.14.

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 2240, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.90, top-1 agreement 0.60) 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
L11N2240+1🕒 shadow specimen t0-v0.1	1.105	0.55	1.363	1.67	“ initially”“ earlier”“ basic”“ initial”“ first”“ especially”“ early”“ already”
L11N2240-1 t0-v0.1	0.903	0.60	1.110	1.36	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.