INDEX

Explanations

references to sports teams, particularly the Royals

oai_token-act-pair · gpt-3.5-turbo

mentions of specific sports teams, particularly the Royals

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

tox

-0.72

cs

-0.71

dot

-0.68

itz

-0.66

dec

-0.65

Xin

-0.64

Buffer

-0.64

pc

-0.64

Cap

-0.63

CT

-0.63

POSITIVE LOGITS

 Royals

3.92

 Orioles

1.36

 Mariners

1.33

 Rockies

1.32

 Chiefs

1.30

Roy

1.27

Kansas

1.25

 Hornets

1.22

 Padres

1.21

 Mets

1.19

Activations Density 0.023%

references to sports teams, particularly the Royals

mentions of specific sports teams, particularly the Royals

No Comments

No Known Activations