INDEX

Explanations

Keywords related to sports teams, specifically Borussia Dortmund and Juventus

oai_token-act-pair · gpt-3.5-turbo

mentions of football clubs, particularly Borussia Dortmund and Juventus

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

orses

-0.72

ONY

-0.71

arters

-0.68

awk

-0.68

Carter

-0.67

wcsstore

-0.67

ogy

-0.66

gemony

-0.65

Psy

-0.64

pha

-0.63

POSITIVE LOGITS

 Dortmund

1.32

 Munich

1.03

liga

1.00

 Bundesliga

0.96

 Guardiola

0.87

 Bayern

0.87

 Juventus

0.84

 MÃ¼

0.78

 Ajax

0.76

celona

0.75

Activations Density 0.017%

Keywords related to sports teams, specifically Borussia Dortmund and Juventus

mentions of football clubs, particularly Borussia Dortmund and Juventus

No Comments

No Known Activations