INDEX

Explanations

phrases related to post-event scenarios or aftermath

oai_token-act-pair · gpt-3.5-turbo

references to post-war or post-event contexts

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

 Cohn

-0.76

 Rivals

-0.76

 Norris

-0.68

onga

-0.66

 Larson

-0.66

 Ukrain

-0.66

Zup

-0.65

 Sieg

-0.62

 Burr

-0.62

 Barron

-0.62

POSITIVE LOGITS

processing

1.09

season

1.06

social

1.02

mortem

0.99

condition

0.99

trial

0.96

release

0.95

design

0.95

regulation

0.95

processor

0.94

Activations Density 0.027%

phrases related to post-event scenarios or aftermath

references to post-war or post-event contexts

No Comments

No Known Activations