INDEX

Explanations

dialogue punctuation, specifically question marks and quotation marks

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly periods and commas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 confir

-0.91

jri

-0.88

ividual

-0.86

 proport

-0.82

 reconc

-0.81

 undermin

-0.81

regate

-0.80

 commer

-0.80

 tremend

-0.78

 volunte

-0.78

POSITIVE LOGITS

 Slowly

1.45

 Suddenly

1.39

 Turning

1.35

 Damn

1.28

 Something

1.26

 Somehow

1.26

 Tears

1.25

 Looking

1.24

 Then

1.23

 Seeing

1.23

Activations Density 0.117%

dialogue punctuation, specifically question marks and quotation marks

punctuation marks, particularly periods and commas

No Comments

No Known Activations