INDEX

Explanations

phrases that indicate rankings or evaluations, particularly regarding competitions

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

tera

-0.71

aji

-0.71

ãĥ£

-0.70

kamp

-0.68

²¾

-0.64

pled

-0.63

kun

-0.63

dit

-0.63

lication

-0.62

pora

-0.62

POSITIVE LOGITS

 luck

0.78

 Nanto

0.64

 Worst

0.64

 breed

0.58

 hindsight

0.58

 behaved

0.56

 bounce

0.56

 Recon

0.56

 circumstance

0.56

 Romance

0.55

Activations Density 0.075%

phrases that indicate rankings or evaluations, particularly regarding competitions

No Comments

No Known Activations