INDEX

Explanations

terms related to the use of something or utilizing a particular method or technique for a specific purpose

oai_token-act-pair · gpt-3.5-turbo

instances of the word "use" and its variations in the context of applications or methodologies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

watching

-0.72

ERROR

-0.72

Mania

-0.71

Discuss

-0.69

angel

-0.67

involved

-0.66

boxing

-0.66

 enjoying

-0.66

essing

-0.65

thood

-0.65

POSITIVE LOGITS

 techniques

0.94

 sophisticated

0.88

fully

0.87

 recycled

0.85

 proprietary

0.85

 outdated

0.80

 cryptography

0.79

 borrowed

0.79

 unconventional

0.77

 technologies

0.77

Activations Density 0.123%

terms related to the use of something or utilizing a particular method or technique for a specific purpose

instances of the word "use" and its variations in the context of applications or methodologies

No Comments

No Known Activations