INDEX

Explanations

phrases indicating significant improvements or advancements

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

essee

-0.88

Interstitial

-0.76

anyahu

-0.71

entary

-0.66

urally

-0.65

 Beware

-0.62

zza

-0.60

 Sins

-0.60

XM

-0.60

gel

-0.60

POSITIVE LOGITS

frog

1.23

 leaps

0.98

olate

0.85

 forward

0.80

hemer

0.79

roads

0.78

Forward

0.76

ering

0.75

olicy

0.74

arts

0.74

Activations Density 0.017%

phrases indicating significant improvements or advancements

No Comments

No Known Activations