INDEX

Explanations

words related to providing explanations or reasons for phenomena

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

sembly

-0.74

ille

-0.73

Ranked

-0.72

ngth

-0.69

opers

-0.68

ontent

-0.66

emies

-0.64

field

-0.63

 shalt

-0.62

ches

-0.62

POSITIVE LOGITS

why

1.36

WHY

1.13

why

1.09

 discrepancies

0.87

how

0.86

Origin

0.82

 explanations

0.81

 inconsistencies

0.80

 disapp

0.77

ĸļ

0.74

Activations Density 0.038%

words related to providing explanations or reasons for phenomena

No Comments

No Known Activations