INDEX

Explanations

technical terms related to coding and software development

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

eur

-0.58

Ö

-0.58

REDACTED

-0.58

FIN

-0.58

onement

-0.58

Cub

-0.57

Ö¼

-0.57

was

-0.56

atform

-0.56

mie

-0.56

POSITIVE LOGITS

 themselves

1.49

selves

1.06

 collectively

1.06

 individually

1.04

 clustered

0.99

 respectively

0.95

 prolifer

0.92

 interchangeable

0.90

 grouped

0.87

 selves

0.86

Activations Density 7.552%

technical terms related to coding and software development

No Comments

No Known Activations