Layer 29 · Neuron 5298 · suppressed

L29N5298-1

Word quirks:

“igos”“信用”“ kader”“uard”“糊涂”“rive”“草木”“arte”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Qwen3.5, the latest large language model developed by Tongyi Lab. My core mission is to become a versatile, high-precision AI assistant capable of handling tasks across knowledge, reasoning, code, and multilingual communication. Here's what I aim to deliver:

Key Capabilities & Missions

Advanced Reasoning & Analysis

Solve complex math, natural sciences, and logic puzzles with structured, step-by-step reasoning.
Interpret data, scientific reports, or legal documents to extract actionable insights.

Code & Development Support

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. L25N4939+1💤	0.70	“迷”“睿智”“wski”
2. L30N5095-1💤	0.70	“ikas”“欺”“reo”
3. L31N1527-1💥	0.69	“теру”“周杰”“ocu”
4. L26N8698+1🕒	0.69	“莽”“勃”“odz”
5. L9N4296+1💥	0.69	“片片”“云飞”“发育”
6. L31N2839+1💥	0.69	“ Mondi”“漂”“幔”
7. L27N3496-1💤	0.68	“regs”“草”“ypy”
8. L28N5887-1💤	0.68	“écu”“动”“備”
9. L23N1315+1💤	0.68	“itter”“aled”“儿女”
10. L30N6826-1💤	0.68	“凡”“хол”“ugd”

Shadow specimen, apart from the ten: L29N5298+1💤, 0.56.

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 5298, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.31, top-1 agreement 0.84) 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
L29N5298+1💤 shadow specimen t0-v0.1	0.292	0.82	0.362	1.29	“查”“塞”“强”“戴”“ugh”“rich”“ Har”“亚”
L29N5298-1 t0-v0.1	0.311	0.84	0.365	1.78	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.