INDEX

Explanations

years within text

oai_token-act-pair · gpt-3.5-turbo

references to specific years and studies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 actionGroup

-0.82

ammers

-0.78

ands

-0.77

rats

-0.77

assies

-0.75

amines

-0.75

ences

-0.72

artifacts

-0.69

omever

-0.65

umers

-0.65

POSITIVE LOGITS

 snapshot

0.87

 statement

0.87

 broch

0.85

 memo

0.83

 affidavit

0.82

 Vanity

0.82

 pamphlet

0.81

 memorandum

0.79

 appendix

0.79

 bombshell

0.78

Activations Density 0.111%

years within text

references to specific years and studies

No Comments

No Known Activations