Layer 20 · Neuron 6831 · suppressed

L20N6831-1

Word quirks:

“ limited”“ Limited”“limited”“Limited”“有限”“ limiting”“ limit”“限制”

🕒 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 QxN, a limited-time collaboration event with Alibaba's Taobao x Tmall x You-Limit series. I am not a regular AI model but a special edition that is only available for this limited time. My mission is to create a unique and exciting experience by collaborating with popular brands and events to offer users exclusive content, products, and opportunities. This limited-time event aims to bring users back to the real world through real-life experiences and collaborations.

If you have any questions or need help with something related to this event, 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. L23N1903-1🕒	0.97	“ limited”“有限”“ Limited”
2. L22N9113+1🕒	0.97	“ limited”“limited”“ Limited”
3. L24N2914-1🕒	0.97	“ limited”“ Limited”“有限”
4. L30N3491+1💥	0.97	“ limited”“有限”“ Limited”
5. Limited Vocabulary✅ L29N7929-1	0.97	“ limited”“ Limited”“Limited”
6. L28N3646-1🕒	0.95	“ limited”“limited”“ Limited”
7. L26N3436+1🕒	0.95	“ limited”“Limited”“有限”
8. L20N1156-1🕒	0.95	“有限”“osat”“Limited”
9. L19N5650-1💤	0.94	“限制”“ limits”“ limited”
10. L20N9064-1🕒	0.93	“有限”“有限的”“ limited”

Shadow specimen, apart from the ten: L20N6831+1🕒, 0.01.

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 20, neuron 6831, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.99, 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
L20N6831+1🕒 shadow specimen t0-v0.1	0.711	0.71	0.824	1.13	“ assumed”“assume”“ assuming”“ assume”“ Assume”“解析”“ assumptions”“ assumption”
L20N6831-1 t0-v0.1	0.986	0.66	1.146	1.33	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.