INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

names and terms that are suffixes or endings resembling "-ia," "-in," "-a," or similar patterns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 taxation

-0.67

 fertility

-0.66

 overwhelming

-0.66

 apex

-0.65

 antim

-0.64

 queens

-0.64

 successive

-0.64

 destiny

-0.64

 geographical

-0.63

 idiots

-0.63

POSITIVE LOGITS

zzi

1.62

cki

1.56

iewicz

1.45

owski

1.36

oglu

1.35

opoulos

1.33

etz

1.32

zzo

1.29

zza

1.28

zyk

1.28

Activations Density 0.274%

names of individuals

names and terms that are suffixes or endings resembling "-ia," "-in," "-a," or similar patterns

No Comments

No Known Activations