Layer 25 · Neuron 856 · boosted

L25N856+1

Word quirks:

“ accepting”“接受”“ acceptance”“ Accept”“ accept”“ accepts”“Accept”“ accepted”

🕒 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 Qwen 9 (also known as C9), the latest version of the Qwen series of large-scale accepting and understanding models developed by Alibaba's AI Research team. My mission is to accept and understand complex instructions, provide high-quality responses across diverse domains, and support users in completing tasks with precision and efficiency.

Key Goals:

Accept and Understand Complex Instructions: I can accept and understand complex instructions, such as multi-step reasoning, data analysis, and content creation.
Provide High-Quality Responses: I aim to provide accurate, detailed, and well

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. L21N8774-1🕒	1.00	“接受”“Accept”“ accepting”
2. L23N3866-1🕒	1.00	“ Accept”“ acceptance”“ accept”
3. L21N3582+1🕒	1.00	“接受”“ accepts”“ accepting”
4. L25N3386+1💥	1.00	“接受”“accept”“ Accept”
5. L23N6530+1🕒	0.99	“ accepted”“ accept”“ Accept”
6. L25N4550-1🕒	0.99	“ acceptance”“接受”“Accept”
7. L23N4858-1🕒	0.99	“接受”“ accept”“ accepted”
8. L22N5919-1🕒	0.99	“ Accept”“ acceptance”“ accept”
9. L27N4347-1🕒	0.98	“ accept”“ accepting”“accept”
10. L22N3468-1💤	0.98	“ Accept”“ accepts”“accept”

Shadow specimen, apart from the ten: L25N856-1🕒, 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 25, neuron 856, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.68, top-1 agreement 0.76) 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
L25N856+1 t0-v0.1	0.684	0.76	0.751	1.33	Top boosted tokens ↑
L25N856-1🕒 shadow specimen t0-v0.1	0.684	0.74	0.603	1.32	“oyer”“ bombe”“ gen”“ Campe”“评估”“形”“ammable”“抗原”

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.