INDEX

Explanations

references to "people" and expressions of collective opinion or sentiment

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

catentry

-0.85

Operation

-0.76

BALL

-0.75

inventoryQuantity

-0.72

PET

-0.68

BuyableInstoreAndOnline

-0.66

NI

-0.65

CHECK

-0.63

âĢ¢âĢ¢âĢ¢âĢ¢

-0.63

POSE

-0.63

POSITIVE LOGITS

folk

1.03

liest

0.97

 entrusted

0.95

 closest

0.94

who

0.88

 composing

0.86

 inhab

0.86

behind

0.83

 nearest

0.83

ourced

0.82

Activations Density 0.040%

references to "people" and expressions of collective opinion or sentiment

No Comments

No Known Activations