INDEX

Explanations

mentions of pundits or commentators

oai_token-act-pair · gpt-3.5-turbo

references to political commentators and pundits

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

urat

-0.72

ĸļ

-0.68

history

-0.66

origin

-0.65

angan

-0.64

alien

-0.64

arcity

-0.63

 stealth

-0.63

mine

-0.62

 fertility

-0.61

POSITIVE LOGITS

 pundits

1.08

 pund

1.05

 commentator

0.93

 commentators

0.88

caster

0.83

 columnist

0.80

odox

0.77

 laureate

0.73

ĳå£«

0.70

sonian

0.70

Activations Density 0.018%

mentions of pundits or commentators

references to political commentators and pundits

No Comments

No Known Activations