INDEX

Explanations

phrases related to grouping or categorization

oai_token-act-pair · gpt-3.5-turbo

references to group dynamics and collaboration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

iron

-0.65

atos

-0.64

tips

-0.61

ãĥ¼ãĥ«

-0.60

Gear

-0.59

ervatives

-0.58

WARN

-0.58

 guid

-0.55

ogenic

-0.54

 Wilde

-0.54

POSITIVE LOGITS

multiple

1.58

 multiple

1.53

 Multiple

1.39

Multiple

1.37

multi

1.32

 single

1.17

 simultaneous

1.16

 multipl

1.16

 multi

1.14

 overlapping

1.13

Activations Density 0.620%

phrases related to grouping or categorization

references to group dynamics and collaboration

No Comments

No Known Activations