INDEX

Explanations

references to titles or headings in a document

oai_token-act-pair · gpt-3.5-turbo

references to titles, particularly in various contexts such as movies, shows, or episodes

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

ramid

-0.71

gm

-0.66

agan

-0.64

verages

-0.64

MQ

-0.63

ters

-0.60

ratom

-0.59

istani

-0.59

akia

-0.59

 adjourn

-0.59

POSITIVE LOGITS

plate

0.99

holder

0.94

plates

0.92

holders

0.91

card

0.90

 CARD

0.85

 sponsor

0.84

pad

0.84

manship

0.82

knife

0.79

Activations Density 0.043%

references to titles or headings in a document

references to titles, particularly in various contexts such as movies, shows, or episodes

No Comments

No Known Activations