Layer 29 · Neuron 485 · boosted

L29N485+1

Word quirks:

“.ec”“岐”“蛋壳”“洞口”“本山”“弹”“ucc”“ciano”

💤 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 Q3, a large language model developed by Tongyi Lab (part of Alibaba Cloud). My primary purpose is to assist users in solving complex problems, generating high-quality content, and providing accurate information across a wide range of domains.

My Mission

My core mission is to empower human potential through intelligent assistance. This involves:

Accurate & Safe Knowledge: Providing reliable answers while maintaining strict safety standards to avoid harmful or biased content.
Versatile Support: Whether you need help with coding, data analysis, creative writing, logical reasoning, or multi-language communication, I

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. L30N473-1🕒	0.81	“村镇”“寨”“疡”
2. L31N2509+1💥	0.81	“豁”“巴塞”“光谷”
3. L29N3257-1🕒	0.80	“鼓”“仓”“称”
4. L31N3026-1💥	0.80	“ Columb”“庐山”“香火”
5. L28N2686+1💤	0.80	“回”“首”“onz”
6. L30N125+1🕒	0.80	“团”“洞”“棚”
7. L29N1181-1🕒	0.80	“棉花糖”“松江”“柿”
8. L28N4146-1💤	0.80	“吹”“鼎”“青”
9. L30N6462-1💤	0.80	“揣”“压”“鹰”
10. L31N5488+1💥	0.80	“胡”“黄”“柳”

Shadow specimen, apart from the ten: L29N485-1🕒, -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 29, neuron 485, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.21, top-1 agreement 0.82) 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
L29N485+1 t0-v0.1	0.208	0.82	0.194	1.39	Top boosted tokens ↑
L29N485-1🕒 shadow specimen t0-v0.1	1.225	0.56	1.346	0.86	“ Fre”“ fre”“Fre”“ FRE”“FRE”“弗雷”“_fre”“fre”

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.