INDEX

Explanations

mentions of soccer-related terms or events

oai_token-act-pair · gpt-3.5-turbo

mentions of soccer

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ysis

-0.74

EStream

-0.72

 Manson

-0.69

BuyableInstoreAndOnline

-0.69

++++++++++++++++

-0.67

 Polaris

-0.67

iannopoulos

-0.65

 Krug

-0.63

 Ples

-0.63

izon

-0.62

POSITIVE LOGITS

oos

1.10

 soccer

0.99

 moms

0.90

soc

0.84

 Soccer

0.82

 stadiums

0.77

 ball

0.76

nell

0.75

mens

0.73

 goalkeeper

0.72

Activations Density 0.019%

mentions of soccer-related terms or events

mentions of soccer

No Comments

No Known Activations