Layer 24 · Neuron 416 · boosted

L24N416+1

Word quirks:

“米”“米的”“府”“ VM”“ Meter”“ Michaels”“金色”“meter”

💤 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 Mi Mi, a large-scale MI MI (Multi-Model Interface) model trained by the Mi Mi team. My mission is to be a mission-driven AI assistant that helps users achieve their goals through mission-oriented design.

What I Do:

Mission Design: I help users define clear, actionable missions (e.g., "Design a 10-minute morning routine" or "Create a budget for a 10-year mission").
Step-by-Step Execution: I break missions into logical steps, ensuring each step aligns with the overall goal.
**Resource

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. L31N8659+1💥	0.92	“米”“米的”“ meter”
2. L22N1819+1💤	0.85	“米”“米的”“瓦”
3. L28N2341+1🕒	0.84	“ mil”“米”“ Mil”
4. L23N4817-1🕒	0.83	“米”“ Rede”“ meter”
5. L27N4142-1💤	0.82	“米”“米的”“usi”
6. L26N6751-1🕒	0.78	“ meter”“meter”“-meter”
7. L23N1717-1🕒	0.78	“ measure”“米”“米的”
8. L29N1215+1💤	0.77	“微”“锰”“ mip”
9. L28N6775-1💤	0.77	“米”“米的”“ Met”
10. L30N6991+1🕒	0.77	“米”“미”“美”

Shadow specimen, apart from the ten: L24N416-1🕒, 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 24, neuron 416, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.63, 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
L24N416+1 t0-v0.1	0.634	0.71	0.694	1.31	Top boosted tokens ↑
L24N416-1🕒 shadow specimen t0-v0.1	0.820	0.74	0.919	1.91	“kast”“舍得”“ Marca”“vale”“ المك”“ders”“思想工作”“ajn”

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.