INDEX

Explanations

references to legs in a variety of contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

 Tycoon

-0.79

NEWS

-0.69

eer

-0.69

 retrospective

-0.66

vertis

-0.65

 Flavoring

-0.65

 uncertainties

-0.64

 beverage

-0.63

urally

-0.63

ãĤ©

-0.62

POSITIVE LOGITS

guards

1.07

flex

0.89

spin

0.85

 amput

0.85

lift

0.85

puter

0.84

chair

0.82

pan

0.82

guard

0.82

 Legs

0.81

Activations Density 0.009%

references to legs in a variety of contexts

No Comments

No Known Activations