INDEX

Explanations

terms related to legal standards and evaluations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

tests and quizzes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

IFrame

Link

Negative Logits

idł

-0.54

InstanceState

-0.49

PhysRev

-0.46

ollection

-0.45

enumi

-0.45

indd

-0.43

참고

-0.43

celes

-0.43

abstractmethod

-0.42

 TreeNode

-0.42

POSITIVE LOGITS

 test

2.23

 tests

2.11

 Test

1.91

 Tests

1.83

test

1.75

Test

1.72

Tests

1.66

 TEST

1.63

 TESTS

1.59

tests

1.54

Activations Density 0.550%

terms related to legal standards and evaluations

tests and quizzes

No Known Activations