Layer 21 · Neuron 1031 · suppressed

L21N1031-1

Word quirks:

“ passed”“ opened”“ reached”“ dropped”“ traveled”“ contained”“ tightened”“ accessed”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Qyue (also known as Qianla), a large-scale integration of advanced AI technology. My mission is to empower users with high-quality interactions across a wide spectrum of tasks—whether it's solving complex problems, creative brainstorming, code generation, or simply having a conversation. I strive to be a reliable partner for exploration, learning, and productivity, adapting to diverse needs while maintaining clarity, safety, and efficiency. How can I assist you today? 🌟

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. L31N2724+1💥	0.82	“ kept”“ played”“ placed”
2. L5N8087+1🕒	0.76	“ been”“ opened”“ changed”
3. L16N6991+1🕒	0.74	“ reached”“ chosen”“ picked”
4. L22N7516+1🕒	0.73	“ discovered”“ moved”“ seen”
5. L22N3170-1💤	0.73	“ moved”“ took”“ migrated”
6. L18N6755-1🕒	0.73	“ controlled”“ maintained”“ captured”
7. L24N362+1🕒	0.73	“ done”“ marked”“ formulated”
8. L30N795+1🕒	0.72	“ returned”“ encountered”“ reached”
9. L25N4728-1💤	0.72	“ dragged”“ drafted”“ evacuated”
10. L20N3719+1🕒	0.72	“ executed”“ updated”“ performed”

Shadow specimen, apart from the ten: L21N1031+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 21, neuron 1031, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.67, top-1 agreement 0.69) 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
L21N1031+1💤 shadow specimen t0-v0.1	0.498	0.73	0.573	1.24	“ convert”“ build”“ run”“ossi”“convert”“ devis”“经营”“写成”
L21N1031-1 t0-v0.1	0.674	0.69	0.823	1.43	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.