INDEX

Explanations

mentions of impairment or impaired states

oai_token-act-pair · gpt-3.5-turbo

terms related to impairment or disability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Aval

-0.75

 Tradition

-0.67

 Bundy

-0.66

ILA

-0.65

Chel

-0.65

pun

-0.64

ICAN

-0.64

 Patriarch

-0.63

 Chrys

-0.63

Tik

-0.62

POSITIVE LOGITS

 impaired

1.49

 impair

1.38

 impairment

1.33

 hemor

0.88

ments

0.85

rador

0.81

 seiz

0.79

 cripp

0.78

intosh

0.78

raints

0.78

Activations Density 0.014%

mentions of impairment or impaired states

terms related to impairment or disability

No Comments

No Known Activations