INDEX

Explanations

specific items or details within a larger set of information

oai_token-act-pair · gpt-3.5-turbo

references to the contents of documents or items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

CVE

-0.65

ARP

-0.64

roads

-0.64

 Architects

-0.63

asio

-0.63

capital

-0.62

ron

-0.62

ansson

-0.62

Norm

-0.61

Di

-0.61

POSITIVE LOGITS

 contents

1.11

uggest

0.99

pread

0.95

afety

0.91

odox

0.88

 Contents

0.86

ĸļ

0.85

iveness

0.82

ngth

0.82

matter

0.81

Activations Density 0.009%

specific items or details within a larger set of information

references to the contents of documents or items

No Comments

No Known Activations