INDEX

Explanations

words related to behaviors or ideas that are opposed to social norms or expected conduct

oai_token-act-pair · gpt-3.5-turbo

terms related to antisocial behavior or concepts

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

 Duchess

-0.92

lly

-0.75

 Penet

-0.72

ity

-0.68

È

-0.67

ãģ®æ

-0.66

 Thumbnails

-0.65

 Falls

-0.64

HRC

-0.64

 Sultan

-0.62

POSITIVE LOGITS

pace

1.20

ocial

1.17

uit

1.13

ystem

1.02

creen

1.02

paces

1.02

uits

0.98

peed

0.98

hirt

0.96

leep

0.96

Activations Density 0.054%

words related to behaviors or ideas that are opposed to social norms or expected conduct

terms related to antisocial behavior or concepts

No Comments

No Known Activations