INDEX

Explanations

phrases related to organizations, healthcare, and social causes

oai_token-act-pair · gpt-3.5-turbo

references to fraudulent or deceptive practices in charitable or healthcare contexts

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

ename

-0.86

ilda

-0.82

aman

-0.77

axe

-0.76

pter

-0.74

forcer

-0.74

 Photographer

-0.73

damage

-0.73

pointer

-0.72

word

-0.72

POSITIVE LOGITS

 organizations

1.93

 organisations

1.80

 institutions

1.74

 societies

1.68

 committees

1.63

 churches

1.63

 nonprofits

1.63

 establishments

1.61

 clinics

1.61

 firms

1.59

Activations Density 0.713%

phrases related to organizations, healthcare, and social causes

references to fraudulent or deceptive practices in charitable or healthcare contexts

No Comments

No Known Activations