INDEX

Explanations

phrases related to completing tasks or achieving goals

oai_token-act-pair · gpt-3.5-turbo

references to missions, particularly those labeled by numbers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Ange

-0.74

Column

-0.69

plet

-0.69

poon

-0.68

bart

-0.68

picking

-0.67

ãģį

-0.64

olls

-0.62

cling

-0.62

Robin

-0.62

POSITIVE LOGITS

 Impossible

1.02

 accomplished

0.99

arily

0.86

 statement

0.85

 undertaken

0.84

 mission

0.83

 Directorate

0.83

 missions

0.83

Vie

0.81

Mission

0.80

Activations Density 0.021%

phrases related to completing tasks or achieving goals

references to missions, particularly those labeled by numbers

No Comments

No Known Activations