Layer 30 · Neuron 654 · boosted

L30N654+1

Word quirks:

“芽”“ Psi”“terminal”“addock”“�”“byt”“ترا”“籲”

💤 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 Qwen3.5, the latest large language model developed by Tongyi Lab. My core mission is to empower human creativity and problem-solving by providing accurate, efficient, and context-aware assistance across a wide range of tasks. Here's how I aim to deliver on this mission:

Key Capabilities Aligned with My Mission:

Precision & Depth

Master 100+ languages with nuanced understanding (e.g., idiomatic expressions, cultural context).
Generate high-quality text (code, reports, stories) and analyze complex data (charts,

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. L31N8320+1💥	0.68	“_RT”“ duk”“ตรา”
2. L30N5120+1💤	0.68	“truc”“ustum”“ارف”
3. L26N904-1💤	0.67	“phil”“bogen”“طين”
4. L24N3052-1💤	0.67	“.lex”“草”“屠”
5. L25N7791-1💤	0.66	“aksa”“�”“一株”
6. L28N5511+1💤	0.66	“�”“umbu”“бера”
7. L23N7542-1💤	0.66	“jax”“�”“雨露”
8. L31N5774+1💥	0.66	“巢”“貔貅”“ trien”
9. L29N5240+1🕒	0.66	“たら”“πι”“tee”
10. L28N1805-1💤	0.66	“uple”“elyn”“朵”

Shadow specimen, apart from the ten: L30N654-1💤, 0.40.

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 30, neuron 654, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.23, top-1 agreement 0.86) 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
L30N654+1 t0-v0.1	0.231	0.86	0.288	1.64	Top boosted tokens ↑
L30N654-1💤 shadow specimen t0-v0.1	0.207	0.86	0.217	1.21	“地”“了几”“地使用”“幾”“地的”“几”“age”“ hình”

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.