INDEX

Explanations

phrases related to components or pieces of something

oai_token-act-pair · gpt-3.5-turbo

references to components or parts within a larger system or context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

chie

-0.81

kers

-0.77

fm

-0.72

standing

-0.71

ker

-0.71

ield

-0.71

geon

-0.69

bern

-0.69

FL

-0.69

thur

-0.68

POSITIVE LOGITS

 components

1.19

 component

1.01

 Components

0.99

 parts

0.97

 subsystem

0.90

onents

0.87

 diagrams

0.83

 elements

0.80

 Component

0.79

 additive

0.79

Activations Density 0.011%

phrases related to components or pieces of something

references to components or parts within a larger system or context

No Comments

No Known Activations