Specimen ledger / L27N2030-1

Layer 27 · Neuron 2030 · suppressed

L27N2030-1

Word quirks:

“ minimum”“minimum”“Minimum”“ минимум”“ Minimum”“minimal”“ minimal”“_minimum”

💤 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 Qwen3.1, the large-scale minimum cost of minimum content (mimo) model developed by Tongyi (part of Alibaba Cloud). My core mission is to serve as a versatile, reliable, and intelligent assistant to help users solve real-world problems across diverse scenarios. Here’s how I aim to deliver on this mission:

Key Goals

Minimize Cost, Minimize Content:

Provide minimal but high-quality responses tailored to user needs, avoiding unnecessary verbosity while ensuring accuracy and relevance.
Optimize for **efficiency and

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. L24N4714-1🕒	0.98	“ minimum”“ Minimum”“minimum”
2. L29N5537+1🕒	0.98	“ Minimum”“Minimum”“ minimum”
3. Bare Minimum✅ L19N8701+1	0.97	“ minimum”“minimum”“最低”
4. L29N3386-1🕒	0.97	“ Minimum”“ minimum”“Minimum”
5. L21N9088-1🕒	0.97	“ minimum”“-min”“minimum”
6. L24N6533-1🕒	0.97	“ minimum”“minimum”“ Minimum”
7. L19N5932+1💤	0.96	“ minimum”“ minimal”“最低”
8. L24N250-1💤	0.95	“ minimum”“ minim”“ monitor”
9. L20N8235+1🕒	0.95	“ minimum”“ minimal”“ essentials”
10. L5N5433-1💥	0.95	“ minimum”“ minimal”“ limit”

Shadow specimen, apart from the ten: L27N2030+1🕒, 0.07.

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 27, neuron 2030, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.46, top-1 agreement 0.75) 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
L27N2030+1🕒 shadow specimen t0-v0.1	0.627	0.72	0.697	1.64	“ Maxwell”“麦”“ má”“ peak”“ МА”“ <<<”“马克”“ Peak”
L27N2030-1 t0-v0.1	0.462	0.75	0.476	1.19	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.