INDEX

Explanations

references to scientific research institutions, specifically the National Institutes of Health and related entities

oai_token-act-pair · gpt-3.5-turbo

references to national health and research institutions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

owner

-0.84

tackle

-0.78

jad

-0.76

 owners

-0.70

bats

-0.70

igans

-0.67

mouth

-0.67

zone

-0.67

hillary

-0.67

boarding

-0.67

POSITIVE LOGITS

 Telescope

0.76

uyomi

0.71

 Museum

0.70

 sequencing

0.70

 Medical

0.69

 Institute

0.68

 Authorization

0.68

 Observatory

0.67

aceutical

0.66

 sequ

0.66

Activations Density 0.082%

references to scientific research institutions, specifically the National Institutes of Health and related entities

references to national health and research institutions

No Comments

No Known Activations