INDEX

Explanations

phrases related to news articles and organizations

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

Not in Any Lists

No Comments

Negative Logits

ELY

-0.76

imum

-0.74

nah

-0.71

vim

-0.71

hered

-0.71

onna

-0.70

netflix

-0.70

essa

-0.70

-0.69

alogue

-0.68

POSITIVE LOGITS

 Journal

1.19

 firms

1.07

 bankers

0.95

 analysts

0.88

 Banking

0.88

 banks

0.86

 banker

0.84

 finan

0.83

 banking

0.82

 executives

0.81

Activations Density 0.040%

phrases related to news articles and organizations

No Comments

No Known Activations