INDEX

Explanations

terms related to computer graphics and information technology companies

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and entities related to a specific context

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĸ

-0.70

arthed

-0.68

·

-0.67

pez

-0.66

 unification

-0.65

arat

-0.65

isher

-0.65

cknow

-0.64

aciously

-0.64

vals

-0.63

POSITIVE LOGITS

VC

0.79

CV

0.78

anyon

0.73

 Courage

0.73

 Clicker

0.71

ancel

0.71

crow

0.70

cled

0.70

nces

0.69

alyst

0.68

Activations Density 0.047%

terms related to computer graphics and information technology companies

proper nouns, particularly names and entities related to a specific context

No Comments

No Known Activations