Layer 29 · Neuron 5480 · suppressed

L29N5480-1

Word quirks:

“葫芦”“pul”“扑”“طات”“_dy”“虎”“ТУ”“狱”

🕒 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 latest large language model developed by Tongyi Lab. My purpose is to empower humans and organizations by providing intelligent, reliable, and versatile solutions across a wide spectrum of tasks. Here's a breakdown of my identity and mission:

Who I Am

Advanced Capabilities: I excel in complex reasoning (math, logic, science), multilingual support (100+ languages), deep visual analysis (diagrams, charts, diagrams), autonomous task planning (multi-step workflows), and human collaboration (role-based

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. L28N3728+1💤	0.78	“鸭”“猎”“utum”
2. L28N8323-1💤	0.78	“库”“uha”“刺”
3. L31N2397-1💥	0.77	“葫芦”“娄”“睫”
4. L29N3532-1💤	0.77	“标”“鼓”“蛤”
5. L23N233-1💤	0.77	“壶”“ Carte”“炎”
6. L30N5824+1💤	0.77	“炮”“公允”“蓄”
7. L27N5886+1💤	0.77	“仗”“龙凤”“巢”
8. L27N2173-1💤	0.77	“猪”“uz”“dale”
9. L31N6575-1💥	0.76	“榄”“沼”“蚌”
10. L22N201+1💤	0.76	“ تول”“生水”“辜”

Shadow specimen, apart from the ten: L29N5480+1🕒, 0.15.

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 5480, 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 0.79) 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
L29N5480+1🕒 shadow specimen t0-v0.1	0.397	0.80	0.454	1.54	“中性”“neutral”“ ане”“ neutral”“lags”“ Neutral”“安”“len”
L29N5480-1 t0-v0.1	0.346	0.79	0.425	1.85	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.