INDEX

Explanations

words related to technological protocols

oai_token-act-pair · gpt-3.5-turbo

references to "protocols" in various 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

 Moody

-0.79

ebted

-0.75

gage

-0.73

iris

-0.71

aez

-0.71

 Lowell

-0.69

rolet

-0.68

lev

-0.66

 Emerson

-0.66

odge

-0.66

POSITIVE LOGITS

 protocol

3.74

 protocols

3.15

 Protocol

2.79

 procedure

1.49

prot

1.39

 etiquette

1.22

 regimen

1.20

 procedures

1.19

 handshake

1.13

Prot

1.13

Activations Density 0.016%

words related to technological protocols

references to "protocols" in various contexts

No Comments

No Known Activations