INDEX

Explanations

prepositional phrases describing downward movement

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "the"

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

replace

-0.72

thood

-0.70

qt

-0.70

craft

-0.67

anan

-0.66

mate

-0.64

 namely

-0.63

SPONSORED

-0.62

cÃ©

-0.60

taker

-0.60

POSITIVE LOGITS

 entire

1.32

 remainder

1.18

 proverbial

1.13

 whole

1.11

 entirety

1.08

 same

1.07

 fastest

1.02

 hardest

1.02

 slightest

0.98

 smallest

0.97

Activations Density 0.416%

prepositional phrases describing downward movement

occurrences of the word "the"

No Comments

No Known Activations