INDEX

Explanations

expressions of personal reflection or emotional reasoning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 Meksiku

-0.52

defaultProps

-0.46

匿名使用者

-0.45

eake

-0.45

PreAuthorize

-0.43

ixe

-0.42

 εφ

-0.42

 purpoſe

-0.42

новниш

-0.41

Carriera

-0.41

POSITIVE LOGITS

 Somehow

0.98

Somehow

0.97

 somehow

0.97

なぜか

0.59

 strangely

0.57

 weirdly

0.55

 irgendwie

0.53

なんとなく

0.52

 oddly

0.51

 mysteriously

0.50

Activations Density 0.352%

expressions of personal reflection or emotional reasoning

No Comments

No Known Activations