INDEX

Explanations

The neuron fires on scholarly domain names or technical fields (e.g. “religion,” “cognition,” “neurocognition,” etc.).

oai_token-act-pair · o4-mini

Triggered by @jyhe0408

scientific and medical terms

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

academic or scientific terms related to various disciplines like religion, cognition, and anatomy.

oai_token-act-pair · claude-3-7-sonnet-20250219

Triggered by @neilrathi

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_22/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 various

-2.00

At

-1.91

 numerous

-1.88

･･

-1.87

What

-1.85

Why

-1.84

 different

-1.83

 encompassing

-1.82

－－

-1.79

POSITIVE LOGITS

 Aunque

1.97

 Además

1.88

</b>

1.81



1.76

WAY

1.69

 Algunas

1.67

 carrot

1.63

’

1.60

เที่ยว

1.60

 dritte

1.60

Activations Density 0.157%

The neuron fires on scholarly domain names or technical fields (e.g. “religion,” “cognition,” “neurocognition,” etc.).

scientific and medical terms

academic or scientific terms related to various disciplines like religion, cognition, and anatomy.

No Known Activations

The neuron fires on scholarly domain names or technical fields (e.g. “religion,” “cognition,” “neurocognition,” etc.).

scientific and medical terms

academic or scientific terms related to various disciplines like religion, cognition, and anatomy.

No Known Activations