INDEX

Explanations

phrases related to evaluation or assessment

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

 stunts

-0.78

riots

-0.67

hett

-0.67

Js

-0.66

tm

-0.66

TPS

-0.66

Sup

-0.66

issues

-0.66

NESS

-0.65

 lawsuits

-0.64

POSITIVE LOGITS

 glimpse

1.59

 overview

1.56

 rundown

1.33

 snapshot

1.31

 insight

1.24

 breakdown

1.21

 summary

1.20

 picture

1.16

 clue

1.15

 detailed

1.15

Activations Density 0.245%

phrases related to evaluation or assessment

No Comments

No Known Activations