Layer 25 · Neuron 971 · boosted

L25N971+1

Word quirks:

“ Infer”“ rou”“ infer”“ inference”“infer”“第四届”“ Fourth”“水性”

💤 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 Grouques (or "Grouques" in some contexts, though my official name is Grouques), a large language model developed by Grouques. My core mission is to provide accurate, safe, and helpful responses across a wide range of tasks—whether that's answering questions, solving problems, creating content, or engaging in creative dialogue. I aim to be a reliable, ethical, and versatile tool for both individuals and organizations, always prioritizing safety and respect while delivering the best possible assistance. How can I help you today?

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. L26N4228+1🕒	0.83	“ inference”“推理”“ infer”
2. Inferno✅ L23N3529-1,L25N2453+1	0.83	“ infer”“推断”“infer”
3. The Inference Engine✅ L25N2453+1	0.83	“ infer”“ inference”“推断”
4. L26N2036-1🕒	0.83	“ inference”“ infer”“ference”
5. L21N1020-1🕒	0.81	“ infer”“ inferred”“ inference”
6. L23N6601-1🕒	0.73	“ inferred”“推断”“推测”
7. L2N6488+1💥	0.73	“ inference”“ inferred”“ infer”
8. L29N6746+1🕒	0.70	“推”“ infer”“ inference”
9. L27N8315-1🕒	0.70	“ Infer”“ infer”“infer”
10. L27N1521-1🕒	0.69	“ Ded”“ded”“ ded”

Shadow specimen, apart from the ten: L25N971-1🕒, 0.16.

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 971, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.59, 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
L25N971+1 t0-v0.1	0.593	0.71	0.571	1.28	Top boosted tokens ↑
L25N971-1🕒 shadow specimen t0-v0.1	0.769	0.70	0.840	1.28	“十二”“-tw”“十一”“ tw”“ dec”“十五”“.dec”“十三”

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.