INDEX

Explanations

technical terms and code elements related to programming

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

 ModelExpression

-1.13

 ujednoznacz

-0.98

 transfieras

-0.97

twimg

-0.93

 autorytatywna

-0.86

 AssemblyCulture

-0.85

 resourceCulture

-0.84

tanleria

-0.84

Autoritní

-0.84

MLLoader

-0.84

POSITIVE LOGITS

↵

0.43

Main

0.43

0.36

↵↵

0.36

 Main

0.34

Setup

0.33

0.32

 main

0.32

Activations Density 0.003%

technical terms and code elements related to programming

No Comments

No Known Activations