INDEX

Explanations

references to the release of reports, studies, statements, polls, and other public materials

oai_token-act-pair · gpt-3.5-turbo

references to the release of documents, reports, or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 befriend

-0.74

Joined

-0.70

SHIP

-0.70

uncture

-0.65

Gad

-0.63

mith

-0.62

het

-0.61

Administ

-0.61

agine

-0.60

mentation

-0.59

POSITIVE LOGITS

 statement

0.81

 verdict

0.81

 teaser

0.77

 damning

0.77

 manifesto

0.76

 roadmap

0.75

 transcripts

0.75

 catalogue

0.75

 censored

0.75

 definitive

0.74

Activations Density 0.285%

references to the release of reports, studies, statements, polls, and other public materials

references to the release of documents, reports, or statements

No Comments

No Known Activations