INDEX

Explanations

various positive attributes or descriptions related to music, particularly in the genres of heavy metal, blues, and dubstep

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ificantly

-0.73

ModLoader

-0.65

okingly

-0.65

ameless

-0.63

versely

-0.62

ependent

-0.62

TPPStreamerBot

-0.58

cially

-0.58

oubted

-0.57

ateful

-0.57

POSITIVE LOGITS

(>

0.96

vironments

0.92

smanship

0.90

 imaginable

0.87

(<

0.86

 awaits

0.79

 syndrome

0.74

isms

0.72

ounters

0.72

izen

0.71

Activations Density 6.192%

various positive attributes or descriptions related to music, particularly in the genres of heavy metal, blues, and dubstep

No Comments

No Known Activations