INDEX

Explanations

the word "who" followed by a name or pronoun

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize a sense of shared experiences or collective observations about individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Delicious

-0.83

ãĥĥãĥī

-0.71

 Bound

-0.67

MSN

-0.64

âĺħ

-0.64

PokÃ©

-0.64

 Lions

-0.63

MER

-0.62

 Pioneer

-0.62

 Mast

-0.62

POSITIVE LOGITS

oping

1.00

oped

0.99

soever

0.90

ikk

0.89

ppers

0.88

 accompanies

0.86

ever

0.85

 preceded

0.84

atever

0.80

 attended

0.78

Activations Density 0.114%

the word "who" followed by a name or pronoun

phrases that emphasize a sense of shared experiences or collective observations about individuals

No Comments

No Known Activations