INDEX

Explanations

phrases related to time periods

oai_token-act-pair · gpt-3.5-turbo

references to different historical or social "post" contexts, such as post-war or post-apocalyptic scenarios

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Rivals

-0.80

 Cohn

-0.80

 Burr

-0.72

 Larson

-0.68

 Ital

-0.67

 Qiao

-0.67

Zup

-0.66

 Barron

-0.65

kson

-0.63

 Wanted

-0.63

POSITIVE LOGITS

processing

1.19

mortem

1.12

jud

0.98

level

0.97

regulation

0.96

election

0.96

death

0.95

condition

0.95

social

0.94

centric

0.94

Activations Density 0.023%

phrases related to time periods

references to different historical or social "post" contexts, such as post-war or post-apocalyptic scenarios

No Comments

No Known Activations