INDEX

Explanations

proper names related to Rupert Murdoch

oai_token-act-pair · gpt-3.5-turbo

mentions of Rupert Murdoch and his associations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RNA

-0.82

venge

-0.76

ranged

-0.75

cession

-0.74

activated

-0.73

bos

-0.71

eping

-0.70

¥µ

-0.70

testing

-0.68

anwhile

-0.68

POSITIVE LOGITS

 Murdoch

1.37

mog

0.95

minster

0.87

doms

0.86

emouth

0.81

 tabl

0.79

enthal

0.78

 tabloid

0.78

vertising

0.77

 mogul

0.74

Activations Density 0.020%

proper names related to Rupert Murdoch

mentions of Rupert Murdoch and his associations

No Comments

No Known Activations