INDEX

Explanations

terms related to fundamental principles or concepts

oai_token-act-pair · gpt-3.5-turbo

references to principles or fundamental concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

minster

-0.85

OOL

-0.73

ummer

-0.73

NetMessage

-0.73

quer

-0.70

otton

-0.70

azo

-0.70

inia

-0.70

ammers

-0.69

imar

-0.68

POSITIVE LOGITS

ually

0.98

 principle

0.84

 principles

0.78

ual

0.78

arily

0.75

ciples

0.75

SourceFile

0.73

 contrad

0.73

inguishable

0.70

cipled

0.70

Activations Density 0.016%

terms related to fundamental principles or concepts

references to principles or fundamental concepts

No Comments

No Known Activations