INDEX

Explanations

symbols and punctuation, particularly those related to emphasis or importance

oai_token-act-pair · gpt-4o-mini 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

Not in Any Lists

No Comments

Negative Logits

ctions

-0.75

gers

-0.73

tto

-0.71

viron

-0.69

 Frenzy

-0.69

glers

-0.68

verted

-0.68

chnology

-0.68

nings

-0.67

breaks

-0.64

POSITIVE LOGITS

ÑĮ

1.08

âĶģ

0.91

åĭ

0.90

ican

0.83

ħ

0.81

åİ

0.78

alon

0.77

aterial

0.76

eas

0.74

ç¥ŀ

0.74

Activations Density 0.010%

symbols and punctuation, particularly those related to emphasis or importance

No Comments

No Known Activations