INDEX

Explanations

mentions of political figures or leaders, specifically with the title "Prime Minister"

oai_token-act-pair · gpt-3.5-turbo

mentions of Prime Ministers

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

anwhile

-0.75

ologies

-0.74

anners

-0.71

bars

-0.66

odcast

-0.66

INESS

-0.66

wark

-0.63

ciplinary

-0.63

pel

-0.63

 chalk

-0.63

POSITIVE LOGITS

 Minister

1.09

val

1.00

knit

0.98

Rib

0.85

 Ministers

0.83

 Directive

0.83

 Accountability

0.77

 ministerial

0.75

Sense

0.74

Grid

0.72

Activations Density 0.017%

mentions of political figures or leaders, specifically with the title "Prime Minister"

mentions of Prime Ministers

No Comments

No Known Activations