INDEX

Explanations

dates and numerical information

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly commas

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

iple

-0.59

ocl

-0.52

OND

-0.50

 Crisis

-0.50

ction

-0.50

rage

-0.48

URE

-0.47

oret

-0.47

 âĨ

-0.47

ongevity

-0.47

POSITIVE LOGITS

 respectively

1.23

etc

0.98

 namely

0.88

 thereby

0.87

anwhile

0.85

 culminating

0.84

aka

0.78

 whereas

0.75

thus

0.73

etc

0.73

Activations Density 0.905%

dates and numerical information

punctuation marks, particularly commas

No Comments

No Known Activations