INDEX

Explanations

phrases describing disparities or inequalities in various contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to disproportionate impacts on various groups or communities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

erm

-0.79

hum

-0.76

ension

-0.75

ource

-0.74

erial

-0.70

zyme

-0.70

aul

-0.70

psons

-0.68

uring

-0.68

ince

-0.67

POSITIVE LOGITS

 disproportionately

0.91

 favoring

0.83

 skewed

0.81

 skew

0.75

 disadvantage

0.74

 amounts

0.73

 disproportion

0.73

 disadvant

0.72

 males

0.72

 minorities

0.72

Activations Density 0.083%

phrases describing disparities or inequalities in various contexts

terms related to disproportionate impacts on various groups or communities

No Comments

No Known Activations