Layer 21 · Neuron 8301 · suppressed

L21N8301-1

Word quirks:

“新鲜”“新鲜的”“鲜”“ marks”“ fresh”“访问”“文章”“新鮮”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am Q-Mars, a large-scale foundation-level large language model developed by Xing Xiang (Xingqian) AI. My primary mission is to provide intelligent, accurate, and accessible solutions to users through natural language interactions. I aim to enhance the user experience by offering high-quality responses to questions, creative content, complex problem-solving, and technical analysis, while ensuring safety and accuracy in all interactions.

If you have any specific questions or tasks, feel free to ask! 😊

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. L24N6903+1💥	0.87	“ fresh”“ Fresh”“鲜”
2. L25N7607-1🕒	0.85	“新鲜”“的新鲜”“ fresh”
3. L20N2071+1💤	0.82	“ fresh”“新鲜”“新鲜的”
4. L25N8698-1💥	0.81	“ Fresh”“鲜”“新鲜”
5. L19N8714+1💤	0.79	“ fresh”“新鲜”“Fresh”
6. L24N2036+1💤	0.78	“监”“怪的”“到中”
7. L31N3479-1💥	0.78	“ fresh”“ Fresh”“鲜”
8. L20N5598+1🕒	0.77	“新生”“ fresh”“ variety”
9. L23N5179-1🕒	0.77	“热风”“ breath”“新鲜”
10. L31N9195-1💥	0.76	“新的”“人的”“ new”

Shadow specimen, apart from the ten: L21N8301+1💤, 0.26.

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 21, neuron 8301, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.90, top-1 agreement 0.66) 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
L21N8301+1💤 shadow specimen t0-v0.1	0.757	0.71	0.839	1.29	“ude”“poll”“ pollution”“ completion”“彻底”“美女”“ admir”“ahlian”
L21N8301-1 t0-v0.1	0.899	0.66	1.069	1.39	Top boosted tokens ↑

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.