INDEX

Explanations

words related to computer graphics and rendering

oai_token-act-pair · gpt-3.5-turbo

technical terms related to rendering, fixes, and exploits in software or gaming contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 trave

-0.80

 notor

-0.75

 Palestin

-0.72

aples

-0.72

 millenn

-0.70

 thous

-0.68

 suspic

-0.67

allery

-0.64

 Seym

-0.64

 concess

-0.63

POSITIVE LOGITS

able

1.68

ings

1.49

ables

1.47

ability

1.46

ers

1.22

athon

1.15

nings

1.12

er

1.09

backs

1.09

ably

1.09

Activations Density 0.360%

words related to computer graphics and rendering

technical terms related to rendering, fixes, and exploits in software or gaming contexts

No Comments

No Known Activations