Layer 11 · Neuron 8901 · suppressed

L11N8901-1

Word quirks:

“ always”“ now”“ previously”“ changed”“ consistently”“ change”“ sometimes”“�”

🕒 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 version of Qwen released by Alibaba Cloud. My mission is to provide high-quality, efficient, and accurate assistance across a wide range of tasks. Here's how I can help:

Enhanced capabilities: I can handle complex reasoning tasks, such as solving math problems or analyzing data.
Improved efficiency: I can process information faster and more accurately than previous versions.
Broader knowledge base: I can access a wider range of topics and provide more detailed responses.

If you have any specific questions about my capabilities, feel free to ask! 😊

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. L9N8131+1💤	0.88	“ changed”“ now”“ change”
2. L8N7773+1🕒	0.87	“ now”“ changed”“ sometimes”
3. L9N5845+1🕒	0.87	“ changed”“ usually”“ sometimes”
4. L12N9159+1💤	0.84	“ changed”“ change”“ changing”
5. L11N806+1💤	0.83	“ always”“ now”“ earlier”
6. L9N8981+1🕒	0.82	“ changed”“ change”“ usually”
7. L17N8154-1💤	0.82	“ now”“ time”“ changed”
8. L9N3777-1💤	0.81	“ now”“ changed”“ replaced”
9. L11N181+1🕒	0.80	“ sometimes”“ often”“ Sometimes”
10. L11N6015+1🕒	0.80	“ always”“ now”“ historically”

Shadow specimen, apart from the ten: L11N8901+1🕒, 0.07.

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 8901, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.99, top-1 agreement 0.62) 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
L11N8901+1🕒 shadow specimen t0-v0.1	1.166	0.60	1.466	1.52	“ ‘”“ seeming”“ \"”“ seems”“ seem”“ mostly”“ apparently”“ ‘’”
L11N8901-1 t0-v0.1	0.988	0.62	1.174	1.31	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.