INDEX

Explanations

information related to news agencies and photo sources

oai_token-act-pair · gpt-3.5-turbo

entities, such as locations and individuals, particularly related to news or reports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

apon

-0.73

utical

-0.73

aterasu

-0.72

 appe

-0.71

udi

-0.70

Ń·

-0.69

igi

-0.65

ushi

-0.65

acers

-0.65

ichick

-0.64

POSITIVE LOGITS

 Caption

0.84

WASHINGTON

0.82

*/(

0.80

................................................................

0.73

BBC

0.71

Press

0.69

https

0.68

SAN

0.67

Loading

0.67

 Cosponsors

0.66

Activations Density 0.499%

information related to news agencies and photo sources

entities, such as locations and individuals, particularly related to news or reports

No Comments

No Known Activations