INDEX

Explanations

words related to regulations or guidelines, especially numerical values

oai_token-act-pair · gpt-3.5-turbo

terms related to various forms of entertainment and enrollment

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

GD

-0.74

ItemTracker

-0.71

 crane

-0.65

Downloadha

-0.63

 Mehran

-0.62

 Mara

-0.62

ilver

-0.57

 breath

-0.57

ODUCT

-0.56

 daddy

-0.55

POSITIVE LOGITS

ments

2.24

ment

2.18

mented

1.62

mentation

1.58

mental

1.55

MENTS

1.47

ations

1.34

MENT

1.33

ation

1.17

cies

1.13

Activations Density 0.072%

words related to regulations or guidelines, especially numerical values

terms related to various forms of entertainment and enrollment

No Comments

No Known Activations