INDEX

Explanations

references to official entities or organizations indicated by the letter "U"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 'U' in the text

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Cind

-0.77

 Emin

-0.67

 Noir

-0.65

 ponies

-0.63

 Siber

-0.63

 responsive

-0.60

 lives

-0.59

 jams

-0.58

 practicable

-0.57

 Scarlet

-0.57

POSITIVE LOGITS

PDATED

1.39

pper

1.28

nexpected

1.28

seless

1.27

zbek

1.18

CLA

1.14

pton

1.11

berman

1.09

prising

1.08

pperc

1.06

Activations Density 0.018%

references to official entities or organizations indicated by the letter "U"

occurrences of the letter 'U' in the text

No Comments

No Known Activations