INDEX

Explanations

_mentions or references to "The New Yorker" publication

oai_token-act-pair · gpt-3.5-turbo

mentions of "The New Yorker" publication

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

leading

-0.73

hold

-0.73

smoking

-0.72

etime

-0.71

ETA

-0.70

held

-0.69

road

-0.68

lower

-0.67

cause

-0.67

rams

-0.66

POSITIVE LOGITS

 Yorker

0.97

vier

0.90

 Wire

0.83

 Mellon

0.81

 Magazine

0.77

 magazine

0.76

 Juda

0.75

 Jacket

0.73

 Swim

0.72

opol

0.70

Activations Density 0.010%

_mentions or references to "The New Yorker" publication

mentions of "The New Yorker" publication

No Comments

No Known Activations