INDEX

Explanations

instances of directional or locational phrases

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

istoitu

-0.88

 kasarigan

-0.74

 MainAxisSize

-0.73

CppCodeGen

-0.71

 snippetHide

-0.70

TypedDataSet

-0.66

 tartalomajánló

-0.65

StructEnd

-0.64

PerformLayout

-0.63

ocino

-0.63

POSITIVE LOGITS

the

0.65

AlterField

0.44

0.40

findpost

0.36

 what

0.35

Bar

0.34

 where

0.34

to

0.31

ArgsConstructor

0.31

bar

0.31

Activations Density 0.029%

instances of directional or locational phrases

No Comments

No Known Activations