INDEX

Explanations

phrases related to agreement or disagreement

oai_token-act-pair · gpt-3.5-turbo

punctuation marks and end of sentences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 clandestine

-0.70

 torpedo

-0.68

 maritime

-0.65

 staggered

-0.64

 circuit

-0.62

 worm

-0.62

 bindings

-0.62

 extinguished

-0.61

 explosive

-0.61

 magistrate

-0.61

POSITIVE LOGITS

âĢ

1.91

 âĢ

1.52

ï¸ı

1.28

âĸł

1.26

âĿ

1.22

âľ

1.21

âĢł

1.19

âĸº

1.18

ðŁĳ

1.16

âĸ

1.15

Activations Density 0.378%

phrases related to agreement or disagreement

punctuation marks and end of sentences

No Comments

No Known Activations