INDEX

Explanations

phrases indicating transformation or change in state

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

erity

-0.89

essim

-0.77

intend

-0.72

ysis

-0.71

 dstg

-0.66

 vulner

-0.65

isha

-0.65

iency

-0.63

 tempted

-0.63

ying

-0.62

POSITIVE LOGITS

 infamous

0.93

 famous

0.85

 known

0.82

 synonymous

0.80

 iconic

0.80

illac

0.78

 staples

0.76

 canon

0.74

famous

0.73

ģĸ

0.72

Activations Density 0.050%

phrases indicating transformation or change in state

No Comments

No Known Activations