INDEX

Explanations

sports-related terms, such as player positions and skills

oai_token-act-pair · gpt-3.5-turbo

phrases indicating strong qualifications or attributes of individuals

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DragonMagazine

-0.87

pmwiki

-0.75

 Networks

-0.70

 Restaur

-0.70

aceae

-0.68

dos

-0.67

 Scroll

-0.67

 Patreon

-0.67

 Locations

-0.66

rss

-0.66

POSITIVE LOGITS

 athletic

1.20

 talented

1.19

 athlet

1.18

 versatile

1.17

 underrated

1.15

 capable

1.11

 gifted

1.06

 unstoppable

1.04

 fearless

0.98

 averaging

0.98

Activations Density 0.199%

sports-related terms, such as player positions and skills

phrases indicating strong qualifications or attributes of individuals

No Comments

No Known Activations