INDEX

Explanations

words related to provocation or incitement

oai_token-act-pair · gpt-3.5-turbo

words related to evidence and justification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dive

-0.80

 Index

-0.76

 Bleach

-0.68

 Mage

-0.68

ADS

-0.65

 Frame

-0.64

pid

-0.63

 Closed

-0.62

Frame

-0.62

 Scale

-0.62

POSITIVE LOGITS

 prov

4.33

prov

1.69

 Prov

1.60

 provocation

1.34

Prov

1.31

 PROV

1.30

 provocative

1.22

 provoke

1.16

 surpr

1.14

 provoking

1.06

Activations Density 0.014%

words related to provocation or incitement

words related to evidence and justification

No Comments

No Known Activations