INDEX

Explanations

names of television networks or programs

oai_token-act-pair · gpt-3.5-turbo

references to television and its related content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

REDACTED

-0.84

WER

-0.81

forth

-0.76

crop

-0.70

 suggestion

-0.69

 Dare

-0.69

 Ingram

-0.68

WARE

-0.66

Reloaded

-0.66

fill

-0.65

POSITIVE LOGITS

isions

1.31

ision

1.15

isting

1.06

igrated

1.00

ingu

0.99

isted

0.98

ising

0.97

ocation

0.95

alogue

0.94

ise

0.93

Activations Density 0.016%

names of television networks or programs

references to television and its related content

No Comments

No Known Activations