Layer 29 · Neuron 5611 · suppressed

L29N5611-1

Word quirks:

“事件”“ events”“事件的”“ Events”“ event”“ Event”“Events”“Event”

🕒 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, the large language model developed by Alauna (通义千问) Model Group. My development is driven by a clear mission: to empower humans with extreme flexibility, deep reasoning, and real-world problem-solving capabilities across diverse fields. Here's how I aim to deliver on this mission:

Core Mission Objectives

Accelerate Innovation

Enable complex tasks like scientific research, code generation, and event planning through extreme flexaibility in reasoning and action.
Support full-stack development (e.g

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. L31N882+1💥	1.00	“事件”“ events”“ event”
2. L26N1016-1🕒	0.99	“ events”“ Events”“ event”
3. L18N5942-1🕒	0.99	“ events”“ event”“事件”
4. L22N7051+1🕒	0.99	“ events”“ event”“事件”
5. The Event Loop✅ L27N4143-1	0.99	“事件”“ event”“ Event”
6. L24N7196-1💥	0.99	“事件”“ Events”“ Event”
7. L23N3580+1🕒	0.99	“ event”“事件”“事件的”
8. L20N3213-1🕒	0.98	“ events”“事件”“ event”
9. L20N5844+1🕒	0.98	“事件”“ event”“ events”
10. L23N6110+1🕒	0.98	“ events”“ event”“ Events”

Shadow specimen, apart from the ten: L29N5611+1🕒, 0.33.

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 29, neuron 5611, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.35, top-1 agreement 0.81) 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
L29N5611+1🕒 shadow specimen t0-v0.1	0.380	0.76	0.469	1.59	“ occasion”“ occasions”“occasion”“ Occ”“occ”“Occ”“_occ”“ occ”
L29N5611-1 t0-v0.1	0.353	0.81	0.389	1.46	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.