INDEX

Explanations

phrases related to academic writing and research

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 'A'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Oswald

-0.66

killers

-0.65

 outlets

-0.63

 Emanuel

-0.61

olicy

-0.59

 Abortion

-0.59

Assembly

-0.59

 appointments

-0.58

 ObamaCare

-0.58

alter

-0.58

POSITIVE LOGITS

cknowled

1.65

cknow

1.58

verages

1.30

chie

1.26

chieve

1.23

perture

1.14

UTH

1.11

uctions

1.08

ctions

1.08

roma

1.07

Activations Density 0.122%

phrases related to academic writing and research

occurrences of the letter 'A'

No Comments

No Known Activations