INDEX

Explanations

technical terms and details related to high-tech equipment and projects

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

CHAT

-0.67

 teasp

-0.67

xtap

-0.66

ĺħ

-0.64

 Brief

-0.63

¬¼

-0.62

ãĤ´ãĥ³

-0.61

 entirety

-0.61

ĳå£«

-0.60

cknow

-0.59

POSITIVE LOGITS

 than

1.97

than

1.88

 Than

1.32

erous

0.90

efficient

0.87

 compared

0.77

lier

0.72

 relative

0.65

worldly

0.65

rah

0.63

Activations Density 3.212%

technical terms and details related to high-tech equipment and projects

No Comments

No Known Activations