INDEX

Explanations

government-related terms in a political context

oai_token-act-pair · gpt-3.5-turbo

references to political figures and their actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Ortiz

-0.93

 Santana

-0.93

caliber

-0.89

 jewelry

-0.89

 behaviors

-0.86

BYU

-0.85

é¾įåĸļå£«

-0.84

 Fresno

-0.83

 neighborhood

-0.82

Houston

-0.82

POSITIVE LOGITS

 Labour

1.69

Labour

1.64

 Ukip

1.58

MPs

1.57

 UKIP

1.56

 Parliamentary

1.55

 parliamentary

1.49

 Tories

1.49

colour

1.41

 Parliament

1.40

Activations Density 1.578%

government-related terms in a political context

references to political figures and their actions

No Comments

No Known Activations