INDEX

Explanations

abbreviation "PU" or variations of it

oai_token-act-pair · gpt-3.5-turbo

abbreviations related to institutions or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

heimer

-0.76

itude

-0.74

 Celest

-0.74

 Journey

-0.72

 Heard

-0.71

heid

-0.71

eworld

-0.71

adm

-0.69

 Remain

-0.68

 Esper

-0.68

POSITIVE LOGITS

BLIC

1.60

PDATE

1.32

PU

1.22

yip

1.04

 mosqu

0.94

 millenn

0.90

cci

0.90

pta

0.90

pse

0.89

 hemor

0.89

Activations Density 0.007%

abbreviation "PU" or variations of it

abbreviations related to institutions or organizations

No Comments

No Known Activations