INDEX

Explanations

occurrences of the word "one"

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

zos

-0.74

keepers

-0.72

LESS

-0.68

ocracy

-0.65

cats

-0.62

letters

-0.61

ursed

-0.61

keeper

-0.60

brates

-0.59

IVES

-0.58

POSITIVE LOGITS

 glance

1.15

 behest

0.89

 moment

0.88

 apiece

0.86

 point

0.85

point

0.83

 expense

0.81

end

0.78

 instance

0.78

 level

0.77

Activations Density 0.009%

occurrences of the word "one"

No Comments

No Known Activations