INDEX

Explanations

politicians and their respective positions or affiliations

oai_token-act-pair · gpt-3.5-turbo

references to specific politicians 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 @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 civilisation

-0.86

 Notting

-0.81

 Kinnikuman

-0.73

 realise

-0.72

 Euro

-0.71

 fulfil

-0.71

LTD

-0.70

 decay

-0.70

 Yugoslav

-0.70

 mould

-0.68

POSITIVE LOGITS

osponsors

1.52

ospons

1.45

 filib

1.34

Senate

1.23

 caucus

1.16

Democrats

1.14

 caucuses

1.14

 Senate

1.14

itol

1.12

Republican

1.11

Activations Density 0.424%

politicians and their respective positions or affiliations

references to specific politicians and their affiliations

No Comments

No Known Activations