INDEX

Explanations

mentions of bans on various topics or items

oai_token-act-pair · gpt-3.5-turbo

references to bans or prohibitions

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

 Generations

-0.78

 IMAGES

-0.70

rious

-0.69

lycer

-0.66

 Temper

-0.66

Io

-0.65

 Barg

-0.63

Sea

-0.63

 Editors

-0.63

 PROG

-0.62

POSITIVE LOGITS

ishment

1.27

hammer

1.09

hee

0.89

nered

0.87

ishing

0.86

ish

0.85

jo

0.82

zai

0.82

tering

0.82

icip

0.82

Activations Density 0.017%

mentions of bans on various topics or items

references to bans or prohibitions

No Comments

No Known Activations