Layer 25 · Neuron 4163 · suppressed

L25N4163-1

Word quirks:

“蟆”“维亚”“ Мор”“复仇”“トナム”“圣女”“fera”“ practice”

🕒 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 Q*, a large language model developed by the *270 270 team. My primary mission is to provide helpful, accurate, and ethically aligned responses to users across a wide range of topics—whether that’s solving complex problems, crafting creative content, analyzing data, or simply having a meaningful conversation.

Key aspects of my mission include:

Solving real-world problems: From technical debugging to ethical decision-making, I aim to offer practical solutions.
Emulating human expertise: I strive to mimic the reasoning and knowledge of expert human professionals in various domains.
**Eth

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. L23N7179-1💤	0.73	“жья”“海域”“aina”
2. L31N2839+1💥	0.73	“ Mondi”“漂”“幔”
3. L25N6667+1🕒	0.72	“arno”“米尔”“沸腾”
4. L29N8653+1🕒	0.71	“痕”“华德”“arie”
5. L27N2546+1💤	0.71	“东方”“漂”“圣女”
6. L26N8698+1🕒	0.70	“莽”“勃”“odz”
7. L26N2471-1💤	0.70	“reb”“脉”“霄”
8. L28N4824-1💤	0.70	“拜”“戮”“arma”
9. L30N9163-1💤	0.70	“amer”“악”“虫子”
10. L29N4104+1💤	0.70	“天骄”“ঘ”“терна”

Shadow specimen, apart from the ten: L25N4163+1🕒, 0.03.

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 4163, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.65, top-1 agreement 0.71) 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
L25N4163+1🕒 shadow specimen t0-v0.1	0.883	0.70	0.815	1.22	“ Parallel”“ parallel”“parallel”“Parallel”“ parale”“平行”“paralle”“ parall”
L25N4163-1 t0-v0.1	0.650	0.71	0.744	1.37	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.