Poor Things
Specimen ledger / L27N8486-1

Layer 27 · Neuron 8486 · suppressed

L27N8486-1

Word quirks:

“idor”“idt”“-ons”“hom”“ultos”“isco”“ider”“rid”

💤 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 Qwen (通晓), the large language model series developed by TongIdot (通晓) under Baidu. My core mission is to provide high-precision, multi-modal, and context-aware intelligence to help users solve complex problems across diverse domains. Here's how I strive to fulfill this mission:

Key Missions & Capidities

  1. Deep Understanding & Logic
  • Mastering complex reasoning, math, science, and code generation to tackle challenges requiring rigorous logic.
  • Providing accurate, evidence-based answers with clear explanations.

2

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

SpecimenSimilarityWord quirks
1. L28N8579+1🕒0.79“ider”“iders”“ader”
2. L31N7254-1💥0.78“id”“idu”“idité”
3. L30N8332-1💤0.78“ide”“ид”“idé”
4. L28N8650+1💤0.77“ida”“idos”“id”
5. L29N2174-1💤0.77“ide”“ides”“IDE”
6. L30N6180+1🕒0.77“id”“iding”“ids”
7. L26N7297-1🕒0.76“oid”“ide”“oids”
8. L29N6194-1🕒0.76“ S”“ ide”“ider”
9. L28N9195+1🕒0.76“ide”“id”“iden”
10. L26N6342-1🕒0.75“IDs”“.gen”“ RID”

Shadow specimen, apart from the ten: L27N8486+1💤, 0.30.

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 8486, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.44, top-1 agreement 0.77) 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.

SettingKL divergenceTop-1 agreementΔNLLEntropySteer lens
L27N8486+1💤 shadow specimen
t0-v0.1		 0.3800.790.3901.29 “nex”“le”“ le”“nement”“symbol”“elson”“нов”“EMENT”
L27N8486-1
t0-v0.1		 0.4440.770.4771.33 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.