INDEX

Explanations

phrases indicating difficulty or inability to achieve something

oai_token-act-pair · gpt-3.5-turbo

the phrase "couldn't" in various contexts relating to inability or lack of capacity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 protected

-0.70

 Published

-0.64

 deprecated

-0.64

 Mechdragon

-0.63

 friendly

-0.63

 guided

-0.62

PU

-0.62

 liberated

-0.60

 seeded

-0.59

 populated

-0.58

POSITIVE LOGITS

't

1.67

\'

0.96

ÃŃ

0.95

kered

0.91

 afford

0.89

ayan

0.86

adian

0.85

ÃĥÃĤ

0.85

itial

0.82

nel

0.82

Activations Density 0.013%

phrases indicating difficulty or inability to achieve something

the phrase "couldn't" in various contexts relating to inability or lack of capacity

No Comments

No Known Activations