INDEX

Explanations

abstract concepts or ideas

oai_token-act-pair · gpt-3.5-turbo

references to abstraction in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ICAN

-0.74

cca

-0.71

oda

-0.69

ACY

-0.66

ods

-0.66

risome

-0.66

attering

-0.65

owder

-0.64

kers

-0.63

tes

-0.62

POSITIVE LOGITS

stract

1.27

 abstract

1.25

Abstract

0.99

 abstraction

0.94

 algebra

0.87

edIn

0.82

 Abstract

0.82

 toile

0.80

ertation

0.79

¥µ

0.78

Activations Density 0.007%

abstract concepts or ideas

references to abstraction in various contexts

No Comments

No Known Activations