INDEX

Explanations

references to fatal incidents

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "fatal."

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

arity

-0.85

annis

-0.79

arta

-0.73

reads

-0.71

orthy

-0.71

XY

-0.70

PT

-0.69

ledge

-0.68

 Bundle

-0.68

arte

-0.66

POSITIVE LOGITS

 fatal

1.23

istically

1.01

istic

0.99

 mosqu

0.87

ãĥ¥

0.85

 Fatal

0.85

 lethal

0.84

 accidents

0.81

 fatally

0.80

 flaw

0.80

Activations Density 0.003%

references to fatal incidents

occurrences of the word "fatal."

No Comments

No Known Activations

references to fatal incidents

occurrences of the word "fatal."

No Comments

No Known Activations