Layer 23 · Neuron 7673 · boosted

L23N7673+1

Word quirks:

“更正”“熏”“改造项目”“更少”“改动”“рово”“ najmniej”“_edit”

💤 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 Qian, a large-scale transformational AI model developed by Canton University of Transformation. My mission is to provide high-quality, accurate, and helpful assistance across a wide range of tasks, including but not limited to:

Answering questions with precise, well-verified information.
Solving problems through logical reasoning and step-by-step guidance.
Creating content such as stories, poems, scripts, or educational materials tailored to your needs.
Analyzing data and providing insights for decision-making.
Learning from feedback to continuously refine my performance.

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. L28N7406+1🕒	0.83	“修改”“改动”“调整”
2. L31N5760-1💥	0.83	“更改”“变更”“改动”
3. L25N8098-1🕒	0.81	“修正”“修改”“重新”
4. L25N1570-1🕒	0.80	“更正”“更生”“更新的”
5. L24N2945+1🕒	0.77	“改变”“更改”“调整”
6. L25N1442-1💤	0.77	“转换”“修改”“变换”
7. L31N8466+1💥	0.77	“改造”“技术改造”“升级改造”
8. L20N6417-1🕒	0.77	“修改”“ modifications”“改造”
9. L29N3094-1🕒	0.75	“转变”“轉變”“修改”
10. L29N6869-1🕒	0.75	“更”“更难”“就更”

Shadow specimen, apart from the ten: L23N7673-1🕒, 0.17.

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 23, neuron 7673, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.64, top-1 agreement 0.70) 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
L23N7673+1 t0-v0.1	0.642	0.70	0.728	1.15	Top boosted tokens ↑
L23N7673-1🕒 shadow specimen t0-v0.1	1.306	0.64	1.421	1.62	“ increased”“增加”“ increase”“ Increased”“ Increase”“Increased”“ increases”“Increase”

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.