INDEX

Explanations

pronouns followed by a verb indicating action or a condition in a condition perspective

oai_token-act-pair · gpt-3.5-turbo 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

 Bound

-0.78

 Anyway

-0.67

Alright

-0.66

PokÃ©

-0.64

Magikarp

-0.63

 Cart

-0.62

 Clover

-0.61

 Compass

-0.60

Rest

-0.60

Sharp

-0.59

POSITIVE LOGITS

oping

1.06

oped

1.02

 voluntarily

0.98

 specialize

0.97

 died

0.96

 underwent

0.93

 participated

0.89

 benefited

0.89

 chronically

0.88

 lived

0.87

Activations Density 0.133%

pronouns followed by a verb indicating action or a condition in a condition perspective

No Comments

No Known Activations