INDEX

Explanations

terms related to Formula 1 racing

oai_token-act-pair · gpt-3.5-turbo

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

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

IFrame

Link

Negative Logits

holder

-0.76

asking

-0.76

gow

-0.74

iannopoulos

-0.72

ulet

-0.71

hat

-0.69

asio

-0.68

asin

-0.68

SPONSORED

-0.67

fully

-0.66

POSITIVE LOGITS

 Drift

0.99

ĩ

0.85

 Formula

0.83

 Renault

0.82

 Britann

0.76

ī

0.73

One

0.72

 racing

0.71

Sport

0.70

XIV

0.67

Activations Density 0.026%

terms related to Formula 1 racing

No Known Activations