INDEX

Explanations

terms related to motion and its representation in media

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

്

-0.68

م

-0.63

KommentareTeilen

-0.54

Autoritní

-0.52

しまう

-0.52

rsiniz

-0.50

atguigu

-0.48

ۦ

-0.47

Aholisi

-0.47

erweise

-0.47

POSITIVE LOGITS

0.72

0.63

0.62

я

0.61

0.55

ه

0.55

ा

0.51

hips

0.48

oğlu

0.47

nesses

0.46

Activations Density 1.576%

terms related to motion and its representation in media

No Comments

No Known Activations