INDEX

Explanations

phrases related to testing and conditions in various scenarios

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

almost

-0.70

 Almost

-0.63

Almost

-0.62

 almost

-0.62

 CallOverrides

-0.61

 kadang

-0.60

 sometimes

-0.60

 GenerationType

-0.59

apparently

-0.58

 onCancelled

-0.58

POSITIVE LOGITS

 properly

1.50

 correctly

1.37

 adequately

1.31

properly

1.20

 correctement

1.20

 möglichst

1.18

 Properly

1.18

 proper

1.18

 adequate

1.18

 appropriately

1.15

Activations Density 0.806%

phrases related to testing and conditions in various scenarios

No Comments

No Known Activations