INDEX

Explanations

words written in a non-English language

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

roleum

-0.73

pez

-0.70

ãģį

-0.68

po

-0.68

fr

-0.67

DP

-0.67

onnaissance

-0.66

oday

-0.66

Els

-0.64

qqa

-0.63

POSITIVE LOGITS

ģĸ

0.77

 cloves

0.75

 glim

0.75

 spices

0.71

 oats

0.70

 lamps

0.66

 extinguished

0.65

IGHTS

0.64

 aven

0.62

 liable

0.62

Activations Density 0.000%

words written in a non-English language

No Comments

No Known Activations