INDEX

Explanations

references to people with specific roles in writing or journalism

oai_token-act-pair · gpt-3.5-turbo

terms related to journalists and their roles as columnists

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

aban

-0.74

inho

-0.73

romy

-0.72

aband

-0.69

illon

-0.69

asonic

-0.69

aeda

-0.68

ypes

-0.66

MSN

-0.66

ulhu

-0.66

POSITIVE LOGITS

 columnist

1.01

 laureate

0.90

itatively

0.90

 writer

0.90

editor

0.82

 bestselling

0.79

 editor

0.78

========

0.75

writ

0.74

column

0.74

Activations Density 0.040%

references to people with specific roles in writing or journalism

terms related to journalists and their roles as columnists

No Comments

No Known Activations