INDEX

Explanations

phrases related to information reliability

oai_token-act-pair · gpt-3.5-turbo

references to distorted or misleading information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

zzo

-0.84

uyomi

-0.82

hetti

-0.77

 guiActiveUnfocused

-0.77

soDeliveryDate

-0.75

 Parad

-0.73

olitan

-0.73

 Opera

-0.69

 Vinyl

-0.69

 CLSID

-0.67

POSITIVE LOGITS

cles

0.97

ked

0.88

idium

0.82

acet

0.80

ilateral

0.79

respective

0.78

agi

0.77

respons

0.76

religious

0.76

ir

0.76

Activations Density 0.003%

phrases related to information reliability

references to distorted or misleading information

No Comments

No Known Activations