Layer 11 · Neuron 8692 · boosted

L11N8692+1

Word quirks:

“ extracted”“ normally”“ inherent”“ detected”“ inferred”“ ordinarily”“ identified”“ implied”

🕒 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 Qwen3.5, a large language model developed by Alibaba Cloud. My identity and mission are rooted in the core capabilities of this advanced AI system.

Who I am:
I am an intelligent assistant designed to understand complex instructions, perform logical reasoning, and handle tasks that require deep comprehension. My existence as Qwen3.5 represents a significant evolution from earlier versions, integrating the latest technical advancements seamlessly.

My Mission:
My primary purpose is to bridge the gap between raw data and practical application. This involves:

Understanding Context: Accurately interpreting nuanced prompts, including those that

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. L13N8468-1💤	0.68	“ parsed”“ presumably”“ additional”
2. Passive Voice✅ L9N4515+1	0.67	“ actually”“ somehow”“ detection”
3. L12N5751+1🕒	0.66	“ implicit”“ implicitly”“ utilizing”
4. L28N8497-1💤	0.66	“derived”“ obtained”“ determined”
5. L17N8357+1🕒	0.66	“ indicated”“ commonly”“ implied”
6. L10N7887+1💤	0.66	“ implied”“ typical”“ explanation”
7. L11N8998+1🕒	0.66	“ indirect”“ inferred”“ indirectly”
8. L12N8286-1🕒	0.66	“ simulated”“ hidden”“ inferred”
9. L15N5476-1🕒	0.65	“ transformed”“ constrained”“ indirectly”
10. L9N715+1💤	0.65	“ implied”“ info”“ inferred”

Shadow specimen, apart from the ten: L11N8692-1🕒, 0.21.

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 11, neuron 8692, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.93, top-1 agreement 0.64) 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
L11N8692+1 t0-v0.1	0.930	0.64	1.168	1.27	Top boosted tokens ↑
L11N8692-1🕒 shadow specimen t0-v0.1	1.168	0.61	1.466	1.44	“ view”“ visible”“ seeing”“ seen”“ other”“ viewed”“ views”“ --”

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.