INDEX

Explanations

references to the name "Ro" followed by a number

oai_token-act-pair · gpt-3.5-turbo

references to the name "Ro" or words beginning with "Ro."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥ¼ãĥĨãĤ£

-0.76

CPC

-0.73

manship

-0.72

lihood

-0.72

ividual

-0.72

 EntityItem

-0.71

 Norn

-0.68

 Beir

-0.68

sson

-0.64

 Birch

-0.63

POSITIVE LOGITS

java

1.03

lette

1.02

oster

1.01

vers

0.98

aches

0.97

zeb

0.94

zzi

0.93

ach

0.93

asted

0.92

zz

0.92

Activations Density 0.009%

references to the name "Ro" followed by a number

references to the name "Ro" or words beginning with "Ro."

No Comments

No Known Activations