INDEX

Explanations

phrases related to news or updates being reported

oai_token-act-pair · gpt-3.5-turbo

phrases that involve reporting or the dissemination of news

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

hement

-0.78

IER

-0.71

kie

-0.68

hesis

-0.68

hesive

-0.68

thumbnails

-0.68

OHN

-0.66

tsky

-0.66

allery

-0.66

alez

-0.66

POSITIVE LOGITS

 impending

1.27

 layoffs

0.94

 imminent

0.93

 executions

0.81

 upcoming

0.78

 whereabouts

0.78

 wrongdoing

0.77

 possible

0.75

how

0.74

fir

0.73

Activations Density 0.135%

phrases related to news or updates being reported

phrases that involve reporting or the dissemination of news

No Comments

No Known Activations