Layer 28 · Neuron 6430 · boosted

L28N6430+1

Word quirks:

“比例”“配合”“ ratios”“ proportions”“的比例”“PRO”“ proportion”“ Norm”

💤 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 in the Qwen series, developed by Tl;P. My purpose is to assist users with a wide range of tasks—providing accurate information, creating professional content, analyzing complex data, and supporting decision-making across industries.

Key Missions:

Proliprimate Intelligence

I aim to bridge the gap between human and machine intelligence by offering natural, intuitive interactions that match or exceed human cognitive capabilities.

Professional & Proliprimate Support

Whether you need to draft a business proposal

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. L19N4943-1🕒	0.93	“ ratios”“比例”“ proportions”
2. L19N6576+1🕒	0.93	“ proportion”“按比例”“比例”
3. L19N5386+1💤	0.93	“ proportion”“ proportional”“比例”
4. L31N6259-1💥	0.93	“比例”“的比例”“ proportion”
5. L24N5449+1🕒	0.93	“ proportion”“比例”“的比例”
6. L21N5108+1🕒	0.93	“比例”“ proportions”“比例为”
7. L22N3821-1🕒	0.92	“比例”“ proportion”“ proportions”
8. Tip-of-the-Tongue✅ L25N2453-1	0.91	“ proportion”“ proportions”“比例”
9. L29N5982+1🕒	0.91	“ proportion”“比例”“ proportional”
10. L19N5635-1💤	0.91	“ proportion”“比例”“的比例”

Shadow specimen, apart from the ten: L28N6430-1💤, 0.20.

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 28, neuron 6430, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.32, 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
L28N6430+1 t0-v0.1	0.320	0.81	0.333	1.33	Top boosted tokens ↑
L28N6430-1💤 shadow specimen t0-v0.1	0.370	0.79	0.419	1.38	“ Morrison”“weet”“ Tha”“拓”“Configuration”“ucion”“ reclam”“经营权”

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.