INDEX

Explanations

words related to general concepts or commonly occurring characteristics

oai_token-act-pair · gpt-3.5-turbo

phrases indicating general trends or common occurrences

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

ÄŁ

-0.86

 Billion

-0.81

 Tycoon

-0.81

yle

-0.78

tein

-0.74

Vaj

-0.72

 Push

-0.71

 Trojan

-0.69

Kut

-0.68

 Tags

-0.68

POSITIVE LOGITS

 regarded

0.85

 speaking

0.84

 exha

0.83

 appreciated

0.82

ensical

0.81

 frowned

0.80

entimes

0.79

 assumed

0.78

 conduc

0.77

 assum

0.77

Activations Density 0.011%

words related to general concepts or commonly occurring characteristics

phrases indicating general trends or common occurrences

No Comments

No Known Activations