INDEX

Explanations

details related to tennis tournaments and players

oai_token-act-pair · gpt-3.5-turbo

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

aneers

-0.87

 Bills

-0.87

âĹ¼

-0.78

[&

-0.78

CAP

-0.77

 smugglers

-0.76

 evac

-0.76

dylib

-0.76

 trooper

-0.75

jriwal

-0.75

POSITIVE LOGITS

 Tennis

1.89

 tennis

1.84

ATP

1.75

bledon

1.65

Wim

1.63

Dj

1.47

Nad

1.39

 Sere

1.37

 tournaments

1.34

 Slam

1.29

Activations Density 0.172%

details related to tennis tournaments and players

No Comments

No Known Activations