Layer 14 · Neuron 7652 · suppressed

L14N7652-1

Word quirks:

“ trying”“ because”“ whenever”“ try”“ when”“ always”“ tries”“ what”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am Qwen, the large language model that powers your interaction with me. My mission is to help users achieve their goals by providing accurate, up-to-date information. Whether it's answering questions, solving problems, or assisting with creative tasks, I strive to be the most helpful assistant possible. Let me know if you need help! 😊

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. L12N8709+1🕒	0.77	“ trying”“ sometimes”“ tries”
2. L11N7519+1🕒	0.77	“ try”“ trying”“ sometimes”
3. L11N9073-1🕒	0.77	“ tried”“ try”“ trying”
4. L12N1841-1🕒	0.76	“ trying”“ somehow”“ tries”
5. L13N8242-1🕒	0.75	“ sometimes”“<\|im_end\|>”“ even”
6. L12N3249+1🕒	0.75	“ trying”“ sometimes”“ related”
7. L12N5557-1🕒	0.74	“ trying”“ tries”“ try”
8. L16N4709+1🕒	0.74	“ actually”“ because”“ trying”
9. L12N5325-1🕒	0.74	“ trying”“ tries”“ try”
10. L10N5349-1🕒	0.73	“ even”“ trying”“ supposedly”

Shadow specimen, apart from the ten: L14N7652+1💤, 0.12.

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 14, neuron 7652, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L14N7652+1💤 shadow specimen t0-v0.1	0.808	0.66	1.042	1.42	“ _____”“ further”“⏎”“ ___”“ __”“ ________”“ again”“ ---”
L14N7652-1 t0-v0.1	1.188	0.58	1.468	1.55	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.