INDEX

Explanations

terms that suggest careful assessment or scrutiny, often in a legal or evaluative context

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 40-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_40/width_16k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.40.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 finally

-0.94

 carefully

-0.94

 easily

-0.89

 finalmente

-0.87

 properly

-0.87

 constantly

-0.87

 quickly

-0.86

 successfully

-0.85

 ultimately

-0.84

 unbedingt

-0.82

POSITIVE LOGITS

tagHelperRunner

0.59

KommentareTeilen

0.53

 enough

0.48

ResumeLayout

0.46

原始内容存档

0.45

地在

0.45

 Numerade

0.45

過ぎる

0.44

 Италијани

0.44

ough

0.43

Activations Density 0.448%

terms that suggest careful assessment or scrutiny, often in a legal or evaluative context

No Comments

No Known Activations