INDEX

Explanations

phrases related to sharing information or updates on social media and online platforms

oai_token-act-pair · gpt-3.5-turbo

instances of the word "posting."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

å§«

-0.81

 Galile

-0.81

estial

-0.75

Ĭ±

-0.73

 Coch

-0.69

vette

-0.66

isin

-0.66

UST

-0.65

 Brill

-0.63

Sco

-0.63

POSITIVE LOGITS

 posting

1.11

 postings

1.10

 posts

1.03

posted

0.88

ulate

0.84

 posted

0.79

doctoral

0.77

aptic

0.77

post

0.76

 selfies

0.76

Activations Density 0.005%

phrases related to sharing information or updates on social media and online platforms

instances of the word "posting."

No Comments

No Known Activations