INDEX

Explanations

links and references identified by their DOI (Digital Object Identifier)

oai_token-act-pair · gpt-3.5-turbo

references to digital object identifiers (DOIs) and bibliographic details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

rete

-0.86

cakes

-0.74

quer

-0.69

ibur

-0.67

 decor

-0.66

reet

-0.66

uci

-0.65

ulz

-0.65

osta

-0.64

iors

-0.63

POSITIVE LOGITS

doi

1.47

DOI

1.17

 PubMed

1.01

ĸļ

1.00

doi

0.95

="#

0.86

romeda

0.83

INGTON

0.83

://

0.80

 CrossRef

0.77

Activations Density 0.012%

links and references identified by their DOI (Digital Object Identifier)

references to digital object identifiers (DOIs) and bibliographic details

No Comments

No Known Activations