INDEX

Explanations

phrases related to additional content or explanation in a text, such as epilogues and flashbacks

oai_token-act-pair · gpt-3.5-turbo

terms related to narrative structures such as prologues and flashbacks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ateurs

-0.91

ateur

-0.83

ello

-0.81

prus

-0.79

ugs

-0.79

athered

-0.79

urrency

-0.78

unte

-0.78

artisan

-0.78

urrencies

-0.78

POSITIVE LOGITS

ogue

1.03

 Prol

0.96

spection

0.85

 Scenes

0.83

 prol

0.77

 transcript

0.75

 Revelations

0.75

 Sequence

0.75

 onwards

0.75

lude

0.74

Activations Density 0.031%

phrases related to additional content or explanation in a text, such as epilogues and flashbacks

terms related to narrative structures such as prologues and flashbacks

No Comments

No Known Activations