INDEX

Explanations

phrases related to recommending or praising things

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 credential

-0.80

 correspond

-0.78

 mosqu

-0.75

 subpoena

-0.75

 concess

-0.74

 subpoen

-0.72

 lobbying

-0.71

 hamstring

-0.70

cffff

-0.70

 displaced

-0.70

POSITIVE LOGITS

Rating

1.65

 Rated

1.40

Overall

1.39

Reviewed

1.29

âĺħâĺħ

1.26

 Helpful

1.23

Review

1.22

Favorite

1.19

Recommended

1.18

Pros

1.18

Activations Density 0.314%

phrases related to recommending or praising things

No Comments

No Known Activations