INDEX

Explanations

phrases containing negative or derogatory terms

oai_token-act-pair · gpt-3.5-turbo

derogatory terms related to people's character or behavior, particularly "asshole"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

neck

-0.87

orders

-0.86

kie

-0.83

estones

-0.79

ching

-0.77

anim

-0.76

bull

-0.76

kered

-0.75

estone

-0.73

å£«

-0.72

POSITIVE LOGITS

mble

0.74

 Catalyst

0.67

APR

0.66

gdala

0.65

 gratification

0.65

 Arbor

0.64

 brackets

0.62

 contrace

0.61

uates

0.61

eering

0.59

Activations Density 0.100%

phrases containing negative or derogatory terms

derogatory terms related to people's character or behavior, particularly "asshole"

No Comments

No Known Activations