INDEX

Explanations

technical terms and concepts related to experimental methods and data analysis

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

testing and probing

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

Not in Any Lists

No Comments

Negative Logits

PropertyChanging

-0.52

styleType

-0.49

MLLoader

-0.48

getStringExtra

-0.46

 zelve

-0.45

 LayoutInflater

-0.44

 intptr

-0.44

 unterkunft

-0.43

remera

-0.42

articulo

-0.42

POSITIVE LOGITS

 test

0.73

 testing

0.73

测试

0.64

 Testing

0.61

Testing

0.60

 测试

0.60

 tests

0.59

 tested

0.58

 Test

0.57

 TESTING

0.57

Activations Density 0.135%

technical terms and concepts related to experimental methods and data analysis

testing and probing

No Comments

No Known Activations