Layer 29 · Neuron 6940 · boosted

L29N6940+1

Word quirks:

“ade”“ trời”“apo”“ソー”“消費”“kre”“caria”“ко”

🕒 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 Q3, a large-scale AIGC (Artificial Intelligence Generated Content) model developed by Tadpoop. My purpose is to provide high-quality, creative, and helpful content across a wide range of topics.

Who I Am

Developer: Tadpoop
Capabilities: I excel in text generation, including writing stories, creating poems, performing scientific reasoning, programming, and more.
Focus: I prioritize accuracy, creativity, and safety in all my interactions.

My Mission

To serve as a reliable and innovative tool for users, helping them:

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. L31N8216-1💥	0.62	“ки”“ki”“cze”
2. L31N8248-1💥	0.62	“retto”“enio”“éos”
3. L30N6294+1💤	0.61	“قي”“кин”“head”
4. L27N4710+1💤	0.61	“科”“ Ko”“Ko”
5. L29N3315-1💤	0.61	“ Ko”“ Karin”“Ng”
6. L28N9154-1💤	0.61	“kon”“apper”“昌”
7. L30N8482+1💤	0.61	“ameran”“бла”“leo”
8. L30N8752+1💤	0.61	“kek”“koa”“anka”
9. L31N2855-1💥	0.61	“دار”“skie”“راع”
10. L25N5603-1🕒	0.61	“ رس”“ADE”“绝”

Shadow specimen, apart from the ten: L29N6940-1🕒, 0.34.

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 6940, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.34, 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
L29N6940+1 t0-v0.1	0.342	0.84	0.412	1.58	Top boosted tokens ↑
L29N6940-1🕒 shadow specimen t0-v0.1	0.348	0.82	0.402	1.52	“ler”“cl”“lea”“kl”“isin”“лей”“LER”“ván”

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.