INDEX

Explanations

references to specific roles or positions

oai_token-act-pair · gpt-3.5-turbo

discussions related to roles in various contexts, particularly in drama, psychology, or societal functions

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

False

-0.76

Latest

-0.73

Ign

-0.71

Sab

-0.68

SB

-0.64

Tea

-0.64

 Rapt

-0.64

Hig

-0.62

BUS

-0.61

Uri

-0.60

POSITIVE LOGITS

 roles

1.23

 role

1.13

role

1.11

 reversal

0.90

playing

0.81

 Role

0.81

ioned

0.81

 incent

0.79

entials

0.75

atel

0.74

Activations Density 0.021%

references to specific roles or positions

discussions related to roles in various contexts, particularly in drama, psychology, or societal functions

No Comments

No Known Activations