INDEX

Explanations

elements of performance evaluation in competitive contexts

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

reality TV shows, elimination, winners

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

memoized

-0.50

//}

-0.47

HasBeenSet

-0.44

 makl

-0.43

чению

-0.43

 ratific

-0.42

 skall

-0.41

 Fermi

-0.41

"]];

-0.40

prepareStatement

-0.39

POSITIVE LOGITS

 contestants

0.89

 elimination

0.85

 contestant

0.82

 eliminated

0.82

elimin

0.81

StructEnd

0.73

 elimin

0.71

 Elimination

0.70

Elimin

0.69

 winner

0.68

Activations Density 0.095%

elements of performance evaluation in competitive contexts

reality TV shows, elimination, winners

No Comments

No Known Activations