INDEX

Explanations

mentions of the award "Vezina Trophy"

oai_token-act-pair · gpt-3.5-turbo

references to specific awards or trophies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

matically

-0.80

matic

-0.77

eering

-0.75

hare

-0.70

hops

-0.69

£ı

-0.67

merce

-0.66

ammy

-0.65

itatively

-0.64

creen

-0.63

POSITIVE LOGITS

Ve

1.35

Ve

1.30

 Veget

0.93

lez

0.89

lette

0.89

ggie

0.86

zza

0.83

Äį

0.79

uve

0.78

ve

0.77

Activations Density 0.005%

mentions of the award "Vezina Trophy"

references to specific awards or trophies

No Comments

No Known Activations