INDEX

Explanations

phrases expressing hypothetical intentions or suggestions

oai_token-act-pair · gpt-4o-mini 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

 Xuan

-0.68

Kag

-0.63

 Chal

-0.62

 Brill

-0.58

è¦ļéĨĴ

-0.57

rylic

-0.56

 Cairo

-0.56

{*

-0.56

rpm

-0.56

 Nile

-0.55

POSITIVE LOGITS

be

0.94

 gladly

0.93

 dearly

0.91

 ideally

0.91

 prefer

0.90

 characterize

0.85

 doubtless

0.80

 ordinarily

0.80

 qualify

0.79

ĸļ

0.79

Activations Density 0.137%

phrases expressing hypothetical intentions or suggestions

No Comments

No Known Activations