INDEX

Explanations

references to news agencies and media organizations

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

profits

-0.70

Â¶

-0.69

grade

-0.62

 totaling

-0.61

abiding

-0.61

 setting

-0.60

 arsenic

-0.60

lycer

-0.59

 langu

-0.59

 authored

-0.59

POSITIVE LOGITS

 IMAGES

0.98

 Images

0.87

Images

0.83

 Wire

0.83

PLIED

0.82

File

0.82

Photos

0.77

Official

0.76

OURCE

0.75

 Photos

0.74

Activations Density 0.078%

references to news agencies and media organizations

No Comments

No Known Activations