Layer 24 · Neuron 8619 · boosted

L24N8619+1

Word quirks:

“系统的”“系统”“系统和”“子系统”“ system”“ SYSTEM”“系统中”“系統的”

💤 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 Qianguang, a large-scale language model developed by the Turingo system. My mission is to provide high-quality, accurate, and helpful information across a wide range of topics—whether it's answering questions, writing text, coding, or solving complex problems. I aim to be an efficient, reliable, and friendly assistant that respects user privacy and adheres to strict safety standards. How can I help you today?

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. L24N7769+1💥	0.97	“系统”“系统的”“系统与”
2. L27N4497+1🕒	0.97	“系统”“系统在”“系统的”
3. L23N6912+1🕒	0.97	“系统”“ system”“系统的”
4. L26N5993-1🕒	0.97	“系统”“系统进行”“系统在”
5. L25N1712-1🕒	0.97	“系统的”“系统”“系统和”
6. L26N2985+1🕒	0.97	“系统”“ systems”“系统与”
7. L24N2152-1💤	0.96	“系统将”“系统”“ System”
8. L23N4004+1🕒	0.96	“系统”“系统的”“ System”
9. L26N8338+1💥	0.96	“系统”“ System”“System”
10. L22N3267-1💤	0.96	“ system”“系统”“系统等”

Shadow specimen, apart from the ten: L24N8619-1💤, 0.07.

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 24, neuron 8619, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.66, top-1 agreement 0.73) 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
L24N8619+1 t0-v0.1	0.661	0.73	0.729	1.28	Top boosted tokens ↑
L24N8619-1💤 shadow specimen t0-v0.1	0.736	0.71	0.776	1.44	“-aut”“sexo”“ nav”“ hall”“sex”“-vertical”“航”“ capacities”

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.