INDEX

Explanations

phrases related to political activism and social justice movements

oai_token-act-pair · gpt-3.5-turbo

phrases indicating demands or actions related to justice and opposition

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ograp

-0.78

 âĨĳ

-0.74

oji

-0.68

availability

-0.65

ocry

-0.61

orem

-0.61

veh

-0.61

Enlarge

-0.61

oultry

-0.61

 Asked

-0.59

POSITIVE LOGITS

enance

1.11

 such

1.07

 this

1.02

 these

1.02

 injustice

0.99

 THIS

0.90

 blindly

0.89

 theirs

0.89

 anything

0.89

 blatant

0.88

Activations Density 0.492%

phrases related to political activism and social justice movements

phrases indicating demands or actions related to justice and opposition

No Known Activations