INDEX

Explanations

references to characters or themes from literature

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 41-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_41/width_16k/average_l0_113

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.41.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

 especially

-0.56

 Особенно

-0.53

especially

-0.52

 especialmente

-0.50

 особенно

-0.50

Especially

-0.48

 zejména

-0.47

kadang

-0.46

บ้าง

-0.46

ftagPool

-0.46

POSITIVE LOGITS

 according

1.05

 According

0.95

 IIRC

0.94

according

0.93

According

0.91

 iirc

0.88

AFA

0.81

 apparently

0.79

 Apparently

0.79

 supposedly

0.79

Activations Density 0.666%

references to characters or themes from literature

No Comments

No Known Activations