Layer 29 · Neuron 2208 · boosted

L29N2208+1

Word quirks:

“锡”“eggi”“aval”“outer”“serir”“台”“ Minggu”“阀”

💤 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 primary mission is to empower users with cutting-edge AI capabilities to tackle real-world challenges—whether that's solving complex problems, creating content, analyzing data, or simply having a natural, helpful conversation.

Key Goals:

Leverage Full-Spectrum Intelligence: From logical reasoning and coding to visual analysis and multi-lingual support, I aim to be a versatile tool for both individuals and organizations.
Simplify Complex Tasks: Whether it's debugging code, planning a trip, or interpreting research

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. L30N6182+1🕒	0.72	“ongs”“三角”“jang”
2. L24N6632-1💤	0.71	“茅”“撤”“庇”
3. L30N4405-1💤	0.71	“针”“开动”“istin”
4. L30N2745-1💤	0.71	“逐”“목”“oller”
5. L31N5221+1💥	0.71	“瓜果”“angka”“牛角”
6. L28N3970+1💤	0.71	“attel”“融”“望”
7. L26N7402+1🕒	0.70	“蛋”“uszcz”“核”
8. L29N9135+1🕒	0.70	“rame”“丸”“叶”
9. L31N1911-1💥	0.70	“藕”“lair”“фра”
10. L29N8549+1🕒	0.70	“策”“社”“宗”

Shadow specimen, apart from the ten: L29N2208-1🕒, 0.09.

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 29, neuron 2208, 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 0.81) 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
L29N2208+1 t0-v0.1	0.230	0.81	0.320	1.30	Top boosted tokens ↑
L29N2208-1🕒 shadow specimen t0-v0.1	1.757	0.42	1.806	1.36	“ purely”“ solely”“ entirely”“ strictly”“纯粹的”“完全是”“纯粹”“ altogether”

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.