INDEX

Explanations

mentions of media and journalism-related terms and actions

oai_token-act-pair · gpt-3.5-turbo

references to media influence and its portrayal or manipulation of events and narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

EStreamFrame

-0.78

tein

-0.78

[|

-0.75

abbage

-0.69

dfx

-0.69

untary

-0.66

 Mobility

-0.65

 Sacrifice

-0.64

ISA

-0.63

Closure

-0.62

POSITIVE LOGITS

 pundits

0.83

 pund

0.79

levision

0.78

 Hannity

0.75

 opinion

0.73

 Editors

0.73

 portraying

0.73

views

0.71

 endorsements

0.70

orters

0.69

Activations Density 0.405%

mentions of media and journalism-related terms and actions

references to media influence and its portrayal or manipulation of events and narratives

No Comments

No Known Activations