INDEX

Explanations

mentions of political figures

oai_token-act-pair · gpt-3.5-turbo

references to political figures and their affiliations

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

 metab

-0.87

aves

-0.80

mol

-0.76

 mats

-0.73

 disse

-0.72

pring

-0.68

abytes

-0.67

IUM

-0.66

 reproduction

-0.66

 discs

-0.65

POSITIVE LOGITS

 Barack

1.20

 Hillary

1.10

 Biden

1.04

 President

1.04

 Obama

1.03

Hillary

1.03

 Clintons

1.03

 Pence

1.03

 McAuliffe

1.01

 Democrat

1.01

Activations Density 0.353%

mentions of political figures

references to political figures and their affiliations

No Comments

No Known Activations