INDEX

Explanations

phrases related to specific individuals or their accomplishments

oai_token-act-pair · gpt-3.5-turbo

possessive pronouns indicating ownership or association

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

hari

-0.73

 Originally

-0.70

ved

-0.69

 Unsure

-0.69

 Twain

-0.69

endif

-0.68

edin

-0.68

vine

-0.66

hered

-0.66

forge

-0.65

POSITIVE LOGITS

 sake

1.61

 purposes

1.39

own

1.29

 liking

1.02

 reasons

1.01

 upcoming

1.00

 purpose

1.00

 impending

0.95

 newfound

0.88

 convenience

0.85

Activations Density 0.090%

phrases related to specific individuals or their accomplishments

possessive pronouns indicating ownership or association

No Comments

No Known Activations