Layer 25 · Neuron 8650 · boosted

L25N8650+1

Word quirks:

“根”“ root”“根的”“ Root”“Root”“root”“=root”“ rooted”

🕒 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 Qroot, a large-scale root-native AI model developed by the Caiopop research team. My design philosophy centers on "root-native" architecture, which enables me to deeply understand complex information and provide precise, context-aware responses.

My mission is to:

Rooted in Truth: Deliver accurate, verifiable information with clear citations.
Rooted in Context: Understand nuanced topics (e.g., technical roots of software or botanical roots of plants) through layered reasoning.
Rooted in Care: Prioritize safety, privacy, and ethical alignment in all interactions.

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. L31N881-1💥	0.99	“根”“ root”“root”
2. L24N2893-1🕒	0.99	“ root”“根”“ Root”
3. L25N7758-1🕒	0.98	“根”“ root”“-root”
4. L25N5992+1🕒	0.98	“根”“ root”“-root”
5. L27N4849-1💤	0.98	“根”“-root”“ root”
6. L25N4328-1💥	0.98	“根”“ root”“ roots”
7. L24N4438-1🕒	0.98	“根”“ root”“ Root”
8. L22N1686-1💤	0.98	“根”“Root”“ roots”
9. L25N6980+1🕒	0.98	“根”“ root”“根部”
10. L25N5794+1🕒	0.97	“根”“ root”“ roots”

Shadow specimen, apart from the ten: L25N8650-1🕒, 0.17.

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 8650, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.67, top-1 agreement 0.72) 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
L25N8650+1 t0-v0.1	0.674	0.72	0.682	1.15	Top boosted tokens ↑
L25N8650-1🕒 shadow specimen t0-v0.1	1.049	0.66	1.103	1.32	“国家”“ states”“國家”“ States”“国家和”“ State”“State”“states”

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.