INDEX

Explanations

abbreviations and punctuation marks

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and sentence endings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

yip

-0.91

 advis

-0.84

 prosec

-0.81

 advoc

-0.77

 challeng

-0.76

 disemb

-0.74

 volunte

-0.73

 averages

-0.71

 negoti

-0.71

 ongoing

-0.70

POSITIVE LOGITS

 Didn

1.46

Was

1.40

Had

1.36

Was

1.35

Had

1.34

 Were

1.30

 Then

1.30

 Afterwards

1.23

 Instead

1.21

 Suddenly

1.21

Activations Density 0.550%

abbreviations and punctuation marks

punctuation marks and sentence endings

No Comments

No Known Activations