INDEX

Explanations

phrases related to technological actions or instructions

oai_token-act-pair · gpt-3.5-turbo

occurrences of a specific symbol or character, likely related to formatting or notational elements in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Franch

-0.91

 snail

-0.78

Sau

-0.75

 Nanto

-0.73

 Isle

-0.73

 Pirate

-0.70

 Postal

-0.69

 Pioneer

-0.68

 Lerner

-0.67

 Moose

-0.66

POSITIVE LOGITS

Ŀ

1.63

¡

1.21

ŀ

1.08

°

1.06

¦

1.00

£

0.98

ī

0.95

©

0.93

§

0.93

ª

0.91

Activations Density 0.282%

phrases related to technological actions or instructions

occurrences of a specific symbol or character, likely related to formatting or notational elements in the text

No Comments

No Known Activations