INDEX

Explanations

octa-, oct-, Stag, TMP

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

acronyms and abbreviations like TMP, OCT, and Stag.

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

Triggered by @neilrathi

The neuron detects uppercase acronyms or initialisms (all-caps words like STAG, TMP, OCT, etc.).

oai_token-act-pair · o4-mini

Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 広場

-3.17

 レスト

-3.09

毟

-3.06

 ヌード

-3.02

簓

-2.97

㩒

-2.80

諼

-2.70

墊

-2.69

 باسلام

-2.67

Ꮿ

-2.63

POSITIVE LOGITS

3.69

2.95

2.92

–

2.78

—

2.67

2.64

2.61

the

2.59

et

2.56

2.52

Activations Density 0.025%

octa-, oct-, Stag, TMP

acronyms and abbreviations like TMP, OCT, and Stag.

The neuron detects uppercase acronyms or initialisms (all-caps words like STAG, TMP, OCT, etc.).

No Known Activations

octa-, oct-, Stag, TMP

acronyms and abbreviations like TMP, OCT, and Stag.

The neuron detects uppercase acronyms or initialisms (all-caps words like STAG, TMP, OCT, etc.).

No Known Activations