INDEX

Explanations

phrases related to competition or competing

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

owe

-0.63

attribute

-0.63

activated

-0.59

relation

-0.59

Adds

-0.57

onics

-0.56

agh

-0.55

many

-0.55

Alz

-0.54

qt

-0.53

POSITIVE LOGITS

 against

1.12

against

0.89

halla

0.86

 Against

0.84

 fiercely

0.83

 competing

0.82

Against

0.77

 unsuccessfully

0.76

 athlet

0.75

 vigorously

0.74

Activations Density 0.019%

phrases related to competition or competing

No Comments

No Known Activations