INDEX

Explanations

HTML closing tags

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-mlp/layer_6/width_131k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.6.hook_mlp_out

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

ArrowToggle

-0.61

 CanadaChoose

-0.60

AutoScaleMode

-0.56

RTRS

-0.46

 unknownFields

-0.42

enumii

-0.41

 ویکی‌پدی

-0.41

 favour

-0.41

Rüyada

-0.40

 defence

-0.40

POSITIVE LOGITS

RenderAtEndOf

0.46

onom

0.44

istem

0.43

imab

0.42

dirent

0.42

ert

0.42

/*

0.41

ient

0.41

jestic

0.41

ONOM

0.40

Activations Density 0.013%

HTML closing tags

No Comments

No Known Activations