INDEX

Explanations

phrases related to protection or listing

oai_token-act-pair · gpt-3.5-turbo

terms related to protection and listing, particularly in legal or regulatory contexts

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

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

 numer

-0.83

 conc

-0.77

 turbo

-0.74

 shar

-0.70

 coefficient

-0.70

 shin

-0.69

 exponent

-0.69

 belie

-0.68

 sten

-0.66

 shake

-0.66

POSITIVE LOGITS

protected

2.54

listed

2.40

protect

1.53

priced

1.47

 unprotected

1.29

registered

1.28

classified

1.23

stable

1.09

 Marines

1.06

encrypted

1.05

Activations Density 0.044%

phrases related to protection or listing

terms related to protection and listing, particularly in legal or regulatory contexts

No Comments

No Known Activations