INDEX

Explanations

names of people, possibly in relation to sports or events

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to sports and athletes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Sabha

-0.74

Assembly

-0.73

 legislatures

-0.63

 Prohibition

-0.63

 Eliot

-0.61

Eag

-0.61

 Hallow

-0.61

 Tablet

-0.60

 Rowling

-0.59

 Assembly

-0.59

POSITIVE LOGITS

ovic

0.93

oglu

0.89

chenko

0.87

Äĩ

0.85

ongo

0.83

ogo

0.82

Lua

0.74

idi

0.73

bye

0.73

ibaba

0.72

Activations Density 0.234%

names of people, possibly in relation to sports or events

proper nouns related to sports and athletes

No Comments

No Known Activations