INDEX

Explanations

mentions of the word "pale" with varying levels of intensity

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

Assembly

-0.73

çīĪ

-0.68

Applications

-0.67

INGTON

-0.65

hips

-0.65

lished

-0.65

ickson

-0.62

sers

-0.61

æ©Ł

-0.60

 arrang

-0.59

POSITIVE LOGITS

olithic

1.57

ozo

1.20

rose

1.01

ogram

1.00

croft

0.93

ocene

0.92

ohyd

0.91

ogen

0.90

ocl

0.90

ographic

0.88

Activations Density 0.020%

mentions of the word "pale" with varying levels of intensity

No Comments

No Known Activations