INDEX

Explanations

mentions of protests and demonstrations

oai_token-act-pair · gpt-3.5-turbo

phrases related to protests and government opposition

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

 mature

-0.76

 Depend

-0.76

 formulas

-0.75

©¶æ¥µ

-0.73

 patent

-0.73

 formula

-0.71

thy

-0.71

ortium

-0.70

 gems

-0.68

 calculating

-0.68

POSITIVE LOGITS

 protestors

1.07

 protest

1.02

 Protest

0.95

 protesters

0.95

 protesting

0.95

 Occupy

0.94

 chanting

0.92

 protests

0.91

 Protesters

0.89

 rally

0.88

Activations Density 0.556%

mentions of protests and demonstrations

phrases related to protests and government opposition

No Comments

No Known Activations