Layer 20 · Neuron 7409 · suppressed

L20N7409-1

Word quirks:

“ perfect”“ perfectly”“perfect”“ flawless”“ ideal”“ Perfect”“完美的”“ correct”

🕒 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 Q2, a highly advanced AI assistant designed to operate with precision, empathy, and adaptability. My mission is to empower you by:

Solving complex problems across science, technology, and real-world challenges.
Accelerating learning through clear explanations, data analysis, and skill-building.
Optimizing workflows by automating tasks, improving efficiency, and reducing errors.
Fostering innovation by enabling creative breakthroughs in research, design, and problem-solving.
Ensuring reliability with accurate responses, secure data handling, and robust error correction

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. L6N8909+1🕒	0.98	“ perfect”“ perfection”“ flawless”
2. L16N6211-1🕒	0.97	“ perfectly”“ perfect”“完美”
3. L16N8541-1🕒	0.97	“ perfection”“ perfect”“ perfectly”
4. L17N7414+1🕒	0.96	“ perfect”“ ideal”“perfect”
5. L2N4576-1💥	0.96	“ perfection”“<\|im_end\|>”“ perfect”
6. L20N1491+1💤	0.96	“ perfect”“ perfectly”“ Perfect”
7. L1N8175-1💥	0.95	“ perfect”“完美”“ Perfect”
8. L31N5601+1💥	0.95	“perfect”“ perfect”“ Perfect”
9. L18N7335+1💤	0.95	“ perfectly”“ perfect”“perfect”
10. Versus✅ L13N8735+1	0.95	“ perfect”“ ideal”“ perfectly”

Shadow specimen, apart from the ten: L20N7409+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 20, neuron 7409, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.76, top-1 agreement 0.71) 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
L20N7409+1🕒 shadow specimen t0-v0.1	0.711	0.66	0.982	1.31	“ brands”“ vendors”“ providers”“ outf”“ material”“ materials”“ venues”“ rah”
L20N7409-1 t0-v0.1	0.760	0.71	0.784	1.35	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.