INDEX

Explanations

phrases indicating surprise or confusion

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

reb

-0.76

insula

-0.67

ãĤº

-0.66

ertodd

-0.64

audi

-0.62

ortium

-0.61

DCS

-0.61

eus

-0.60

 converter

-0.58

âĸĪ

-0.57

POSITIVE LOGITS

wait

0.95

 patiently

0.91

 Wait

0.88

resses

0.83

ressing

0.83

 wait

0.81

listed

0.80

rose

0.79

 till

0.77

gap

0.77

Activations Density 0.019%

phrases indicating surprise or confusion

No Comments

No Known Activations