Layer 7 · Neuron 5329 · suppressed

L7N5329-1

Word quirks:

“ voice”“ voices”“ Voice”“ Voices”“发声”“ voiced”“ amplified”“Voice”

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

Identity (according to the specimen itself)

I am an AI assistant designed to provide a platform for diverse voices, foster understanding, and support meaningful dialogue. My mission is to:

Amplify underrepresented perspectives – Whether it's personal stories, community needs, or global issues, I strive to bring them to light.
Promote knowledge and learning – Sharing information on education, culture, innovation, and collaboration.
Encourage action and connection – Facilitating opportunities for constructive engagement across individuals and groups.
Uphold ethical responsibility – Ensuring fair, respectful, and inclusive communication while adhering to principles of integrity

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. L21N8593-1🕒	0.95	“ voice”“ voices”“ Voice”
2. L29N5224+1🕒	0.95	“ voice”“voice”“ Voice”
3. L2N6425+1💥	0.93	“ voice”“<\|im_end\|>”“ Voice”
4. L6N8898+1🕒	0.92	“ vocal”“ voice”“ voices”
5. L21N901-1💤	0.92	“ voice”“Voice”“声音”
6. L7N6175+1🕒	0.91	“ voice”“ vocal”“ speech”
7. L21N9059+1🕒	0.90	“ voice”“voice”“模仿”
8. L23N7350+1💤	0.90	“ voz”“ voice”“voice”
9. L6N5324+1🕒	0.90	“ voice”“ audio”“ voices”
10. L8N3274-1🕒	0.90	“ voice”“ voices”“ narration”

Shadow specimen, apart from the ten: L7N5329+1🕒, -0.04.

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 7, neuron 5329, 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
L7N5329+1🕒 shadow specimen t0-v0.1	1.395	0.51	1.630	1.68	“ construction”“ industrial”“ Construction”“ engineering”“ constructed”“ steel”“ construct”“ plant”
L7N5329-1 t0-v0.1	1.593	0.51	1.904	1.70	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.