INDEX

Explanations

words related to verification, confirmation, or validation

oai_token-act-pair · gpt-3.5-turbo

concepts related to vision impairment and validation processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

oria

-0.77

ydia

-0.71

 Viet

-0.69

gd

-0.69

riv

-0.69

imb

-0.69

iture

-0.67

 Gamb

-0.67

Ń·

-0.64

 Recreation

-0.63

POSITIVE LOGITS

 blinded

2.93

 blinding

1.73

blind

1.23

 validated

1.21

 masked

1.01

 obscured

0.99

 validate

0.96

 validation

0.93

 distingu

0.83

cookie

0.78

Activations Density 0.023%

words related to verification, confirmation, or validation

concepts related to vision impairment and validation processes

No Comments

No Known Activations