INDEX

Explanations

words related to bias or prejudice, particularly in contexts like news reporting, social networks, and investigations

oai_token-act-pair · gpt-3.5-turbo

phrases related to bias and prejudice in various contexts

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

clamation

-0.79

export

-0.77

isting

-0.76

ODE

-0.71

esc

-0.70

Interstitial

-0.70

ctica

-0.68

PATH

-0.68

sis

-0.68

Money

-0.68

POSITIVE LOGITS

 biases

1.06

 bias

1.05

 prejud

0.96

 favoring

0.86

 toward

0.86

 towards

0.85

 prejudice

0.82

 inference

0.80

 biased

0.80

 detector

0.79

Activations Density 0.023%

words related to bias or prejudice, particularly in contexts like news reporting, social networks, and investigations

phrases related to bias and prejudice in various contexts

No Comments

No Known Activations