INDEX

Explanations

sculptures or sculptural elements within a text

oai_token-act-pair · gpt-3.5-turbo

mentions of sculptures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

FFFF

-0.61

 Canary

-0.61

ITNESS

-0.61

eely

-0.60

 Witnesses

-0.59

ears

-0.59

rah

-0.59

zee

-0.59

Introduced

-0.58

 Outlook

-0.58

POSITIVE LOGITS

 sculpture

1.37

 sculptures

1.29

 statue

0.99

 statues

0.90

 artwork

0.88

 plaza

0.83

 mural

0.82

 paintings

0.81

ĸļ

0.81

 sculpt

0.81

Activations Density 0.011%

sculptures or sculptural elements within a text

mentions of sculptures

No Comments

No Known Activations