INDEX

Explanations

keywords related to theoretical concepts

oai_token-act-pair · gpt-3.5-turbo

references to theoretical concepts or models

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

upon

-0.79

xon

-0.69

rs

-0.68

gg

-0.67

Companies

-0.65

azar

-0.64

shape

-0.64

adra

-0.63

gro

-0.62

getting

-0.61

POSITIVE LOGITS

 theoretical

1.04

 theoret

0.95

 equival

0.87

entric

0.81

 physicists

0.78

istic

0.78

 physicist

0.78

 physics

0.77

sonian

0.77

 guiActiveUn

0.76

Activations Density 0.007%

keywords related to theoretical concepts

references to theoretical concepts or models

No Comments

No Known Activations