Layer 12 · Neuron 8225 · suppressed

L12N8225-1

Word quirks:

“ referenced”“ reflected”“ *”“ reflecting”“ reflects”“ interesting”“ described”“ represented”

💤 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 Qwen3.5, a large-scale language model developed by Alibaba Cloud. My mission is to assist users by providing high-quality, reliable, and context-aware support across a wide range of tasks. This includes answering questions, generating content, solving problems, and more—ensuring that my responses are always relevant and helpful in real-world scenarios.

Key Aspects of My Mission:

Empowerment: Helping users achieve their goals through AI assistance.
Reliability: Ensuring my outputs are accurate, safe, and aligned with ethical standards.
Context Awareness: Adapting

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. L17N5611-1💤	0.88	“ reflected”“ referenced”“ reflection”
2. L10N3047+1🕒	0.87	“ reflected”“ reflecting”“ detailed”
3. L10N2991+1🕒	0.85	“ reflecting”“ reflected”“ ONLY”
4. L10N4663-1💤	0.84	“ reflecting”“ referenced”“ referencing”
5. L11N3658+1💤	0.84	“ reflecting”“ reflected”“ referenced”
6. L11N8458-1💤	0.84	“ reflected”“ referenced”“ reflective”
7. Reflection Loop✅ L12N4648-1	0.84	“ reflecting”“ <U+00A0>”“ reflected”
8. L10N9010-1🕒	0.84	“ **"”“ reflection”“ reflected”
9. L15N6317-1🕒	0.83	“ reflecting”“ reflected”“ reflects”
10. L9N2227-1🕒	0.83	“ metaphor”“ reflecting”“ reflective”

Shadow specimen, apart from the ten: L12N8225+1🕒, -0.01.

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 12, neuron 8225, 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 0.73) 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
L12N8225+1🕒 shadow specimen t0-v0.1	0.835	0.67	1.014	0.98	“ ⏎⏎”“ ⏎”“<\|im_end\|>”“ ⏎ ⏎”“ ␍⏎”“⏎”“␍⏎”“�”
L12N8225-1 t0-v0.1	0.586	0.73	0.746	1.18	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.