INDEX

Explanations

identifiers related to metadata and categorization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

ilon

-0.76

berman

-0.75

agos

-0.68

enthal

-0.67

undai

-0.67

yond

-0.66

rica

-0.64

arios

-0.63

rams

-0.62

unal

-0.62

POSITIVE LOGITS

 Description

1.03

 Tags

1.00

Description

0.99

 Details

0.93

 Reviews

0.93

 Flavoring

0.90

Title

0.90

 Ratings

0.90

 Info

0.86

 Notes

0.86

Activations Density 0.033%

identifiers related to metadata and categorization

No Comments

No Known Activations