INDEX

Explanations

words related to negative attitudes or prejudices, particularly relating to specific groups or characteristics

oai_token-act-pair · gpt-3.5-turbo

terms related to various forms of phobia and intolerance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

MER

-0.85

aver

-0.82

unes

-0.78

upper

-0.76

allo

-0.73

selection

-0.71

mel

-0.71

arist

-0.71

aple

-0.70

selected

-0.70

POSITIVE LOGITS

ophobia

1.31

ophobic

1.22

obia

1.04

ï¸

1.00

ophob

0.99

obic

0.94

osate

0.84

 prejudice

0.81

 bigotry

0.81

 homophobia

0.79

Activations Density 0.012%

words related to negative attitudes or prejudices, particularly relating to specific groups or characteristics

terms related to various forms of phobia and intolerance

No Comments

No Known Activations