INDEX

Explanations

Bayesian, diffusion, variational, evolutionary models

np_acts-logits-general · gemini-2.5-flash-lite

The neuron consistently activates on methodological and framework terms—labels of research designs, models, or technical approaches.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

phrases naming formal technical frameworks or methodologies—compound, often hyphenated academic jargon for models, architectures, algorithms, or approaches.

oai_token-act-pair · gpt-5 Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

वत्ता

0.60

 Monkeys

0.58

 kami

0.56

 ڈپاز

0.56

issantes

0.56

 новых

0.55

 kinderen

0.55

 pommes

0.55

 olduk

0.55

 intenso

0.55

POSITIVE LOGITS

 approach

0.93

 подход

0.88

 Approach

0.76

Approach

0.73

approach

0.73

 abordagem

0.72

 подхода

0.71

 method

0.68

 পদ্ধতিতে

0.68

 methodology

0.67

Activations Density 0.001%

Bayesian, diffusion, variational, evolutionary models

The neuron consistently activates on methodological and framework terms—labels of research designs, models, or technical approaches.

phrases naming formal technical frameworks or methodologies—compound, often hyphenated academic jargon for models, architectures, algorithms, or approaches.

No Comments

No Known Activations

Bayesian, diffusion, variational, evolutionary models

The neuron consistently activates on methodological and framework terms—labels of research designs, models, or technical approaches.

phrases naming formal technical frameworks or methodologies—compound, often hyphenated academic jargon for models, architectures, algorithms, or approaches.

No Comments

No Known Activations