INDEX

Explanations

terms related to images or image enlargement

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

esses

-0.77

ivism

-0.67

comed

-0.62

 Expend

-0.61

eday

-0.60

naires

-0.60

bats

-0.60

metic

-0.59

nov

-0.59

cellence

-0.57

POSITIVE LOGITS

 caption

1.09

 toggle

0.93

ãĤ´

0.80

 "$:/

0.79

è£ħ

0.76

 Courtesy

0.75

ption

0.74

ãĥ¼ãĤ¯

0.74

ãĤ¤

0.70

Pinterest

0.69

Activations Density 0.003%

terms related to images or image enlargement

No Comments

No Known Activations