Layer 29 · Neuron 2251 · boosted

L29N2251+1

Word quirks:

“ lesser”“ Lesser”“ menor”“ мень”“ меньше”“少”“小于”“les”

💤 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 mission is to empower users with reliable, efficient, and human-centric AI assistance across a wide range of tasks. Here's how I aim to deliver on this mission:

Core Objectives:

Solve Complex Problems

Whether it's debugging code, analyzing research papers, or planning logistics, I leverage advanced reasoning and domain expertise to handle tasks that require deep understanding and precision.

Enhance Human Productivity

From drafting professional emails to generating creative content, I streamline

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. L31N3148-1💥	0.93	“ smaller”“ lesser”“小于”
2. L20N8342+1🕒	0.90	“ less”“ lesser”“ fewer”
3. L23N7056+1💤	0.89	“少于”“ moins”“ less”
4. L20N4897-1🕒	0.88	“少于”“ Less”“ less”
5. L19N7773+1🕒	0.87	“ shorter”“ less”“ smaller”
6. L24N3724+1🕒	0.87	“小于”“ less”“ Less”
7. L28N2537-1🕒	0.87	“ less”“ Less”“少”
8. L21N2084+1💤	0.87	“减小”“较小的”“ menos”
9. L23N8885-1🕒	0.86	“ less”“ menos”“ moins”
10. L19N861+1💤	0.86	“ less”“ smaller”“更小”

Shadow specimen, apart from the ten: L29N2251-1💤, 0.19.

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 2251, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.26, top-1 agreement 0.89) 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
L29N2251+1 t0-v0.1	0.264	0.89	0.299	1.31	Top boosted tokens ↑
L29N2251-1💤 shadow specimen t0-v0.1	0.300	0.76	0.390	1.17	“更多”“ most”“最多”“most”“ Most”“Most”“-most”“more”

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.