INDEX

Explanations

references to programming constructs, specifically classes and files in code

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

 BrowserModule

-0.39

Notae

-0.32

<eos>

-0.31

 nuevas

-0.30

 dieux

-0.30

mbggenerated

-0.30

ใหม่

-0.29

Географиясе

-0.29

 mères

-0.28

 épaules

-0.28

POSITIVE LOGITS

DockStyle

0.85

 MyBatis

0.65

MLLoader

0.60

ngdoc

0.59

WebElementEntity

0.59

 queſto

0.57

InputTagHelper

0.57

findpost

0.57

complexContent

0.56

 メンテナ

0.55

Activations Density 0.001%

references to programming constructs, specifically classes and files in code

No Comments

No Known Activations