INDEX

Explanations

information related to publications, scientific papers, and dois

oai_token-act-pair · gpt-3.5-turbo

references and citations in academic writing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 screws

-0.71

 frost

-0.68

clock

-0.68

 sentimental

-0.67

 curtains

-0.66

 bulldo

-0.66

 allot

-0.66

 ponds

-0.65

 eviction

-0.65

 pancakes

-0.65

POSITIVE LOGITS

ournals

1.22

 Retrieved

1.22

doi

1.05

 PLoS

0.93

 Proceedings

0.89

Journal

0.88

 PubMed

0.87

 Supplement

0.84

 Neuroscience

0.83

 Springer

0.83

Activations Density 0.703%

information related to publications, scientific papers, and dois

references and citations in academic writing

No Comments

No Known Activations