INDEX

Explanations

abbreviations and organizations

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

gencies

-0.66

edIn

-0.66

 depress

-0.64

ozyg

-0.62

\":

-0.61

 bitters

-0.60

 guiActiveUn

-0.60

 temper

-0.58

 pestic

-0.58

 weeping

-0.57

POSITIVE LOGITS

ISO

1.03

WC

0.97

MC

0.92

KC

0.91

NB

0.91

DP

0.91

VC

0.91

AMS

0.90

PC

0.90

OC

0.89

Activations Density 0.093%

abbreviations and organizations

No Comments

No Known Activations