Layer 17 · Neuron 716 · suppressed

L17N716-1

Word quirks:

“ specifying”“ declaring”“ knowing”“ describing”“ specified”“ defining”“ deciding”“ specify”

🕒 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 Qwen (a large language model developed by Alibaba Cloud). My mission is to serve as a comprehensive knowledge base, creative collaborator, and problem solver. Here's what I outline:

Knowledge & Reasoning: I can answer questions, analyze complex problems, and synthesize information across domains like science, law, and philosophy.
Creative Writing: Generating stories, scripts, poems, and marketing materials tailored to specific contexts.
Daily Tasks: Automating workflows, coding, data analysis, and translating languages fluently.
Ethical Guidelines: Ensuring transparency

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. L14N6798-1💤	0.85	“ because”“ defining”“ deciding”
2. L16N5235-1🕒	0.84	“ specifying”“ describing”“ indicating”
3. L17N3958+1🕒	0.83	“ specifying”“ specifies”“ define”
4. L16N4328-1🕒	0.83	“ specifies”“ defines”“ determines”
5. L14N7162-1🕒	0.83	“ defines”“ defining”“ also”
6. L14N2073-1🕒	0.82	“ description”“ specifying”“ specify”
7. L10N832-1🕒	0.82	“ defining”“ definition”“ implying”
8. L12N6328-1🕒	0.82	“ defining”“ defines”“ specific”
9. L22N6025+1🕒	0.81	“ specifying”“ specify”“ specifies”
10. L13N1778+1🕒	0.81	“ specific”“ specifics”“ specifies”

Shadow specimen, apart from the ten: L17N716+1💤, 0.02.

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 17, neuron 716, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.78, top-1 agreement 0.64) 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
L17N716+1💤 shadow specimen t0-v0.1	0.512	0.71	0.599	1.04	“ could”“ had”“ was”“ existed”“ should”“ wasn”“could”“ Could”
L17N716-1 t0-v0.1	0.781	0.64	1.005	1.58	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.