INDEX

Explanations

years as expressed in four-digit numbers

oai_token-act-pair · gpt-3.5-turbo

references to the year 2011

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 stret

-0.69

 ether

-0.66

DonaldTrump

-0.63

âĶĢ

-0.63

 escorted

-0.62

ople

-0.58

caster

-0.56

CHO

-0.56

atem

-0.55

achy

-0.54

POSITIVE LOGITS

3.31

2.83

2.77

2.73

2.62

2.58

2.56

2.46

2.44

2.36

Activations Density 0.017%

years as expressed in four-digit numbers

references to the year 2011

No Comments

No Known Activations