INDEX

Explanations

phrases related to sitting back and relaxing

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

Negative Logits

origin

-0.67

catentry

-0.66

Ĥİ

-0.59

 Cantor

-0.58

 Pett

-0.58

Ĵ

-0.57

bringer

-0.57

DeL

-0.57

FER

-0.56

 affiliate

-0.56

POSITIVE LOGITS

 passively

0.96

 awaiting

0.86

 relax

0.85

 smug

0.82

 waiting

0.78

 idle

0.77

 bliss

0.77

 staring

0.75

 comfortably

0.75

 nervously

0.73

Activations Density 0.074%

phrases related to sitting back and relaxing

No Known Activations