INDEX

Explanations

references to images and their associated data types

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

essentiel

-0.50

few

-0.50

SBATCH

-0.49

excellents

-0.47

aucune

-0.47

ONOM

-0.45

Few

-0.45

 voorwaarden

-0.44

Few

-0.44

for

-0.43

POSITIVE LOGITS

 image

1.12

 Image

1.12

Image

1.11

image

1.05

 IMAGE

0.97

getImage

0.96

 imagen

0.94

 getImage

0.93

IMAGE

0.89

imageFile

0.85

Activations Density 0.039%

references to images and their associated data types

No Comments

No Known Activations