INDEX

Explanations

unintelligible strings that do not form coherent words or phrases

oai_token-act-pair · gpt-3.5-turbo

end-of-text markers or invisible delimiters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Pebble

-0.73

 Vanderbilt

-0.72

 cloning

-0.71

 Walmart

-0.71

 Torch

-0.70

 superheroes

-0.68

 scares

-0.68

 Greene

-0.67

 startup

-0.67

 shockingly

-0.66

POSITIVE LOGITS

Ãī

1.20

Ô

1.19

Ã

1.13

Ãį

1.12

Ã©s

1.08

Translation

1.08

 Ã©

1.07

 Ãī

1.04

Ã©t

1.02

Ãª

1.01

Activations Density 0.189%

unintelligible strings that do not form coherent words or phrases

end-of-text markers or invisible delimiters

No Comments

No Known Activations