INDEX

Explanations

years in the 1800s

oai_token-act-pair · gpt-3.5-turbo

historical years

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

aturday

-0.72

category

-0.63

reon

-0.61

lying

-0.59

 whipped

-0.59

elines

-0.59

pse

-0.58

Neigh

-0.58

no

-0.56

Stay

-0.56

POSITIVE LOGITS

 onwards

0.88

 onward

0.80

chev

0.79

å£«

0.73

çļ

0.71

âĶĢâĶĢâĶĢâĶĢ

0.70

 Redux

0.68

-'

0.66

 Ferdinand

0.64

 Edition

0.63

Activations Density 0.070%

years in the 1800s

historical years

No Comments

No Known Activations