INDEX

Explanations

phrases related to research methodologies and search strategies

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

guien

-0.63

 StatelessWidget

-0.61

homonymie

-0.56

unless

-0.55

 AssemblyProduct

-0.54

それは

-0.53

 ujednoznacz

-0.53

 kasarigan

-0.52

Everybody

-0.52

sự

-0.51

POSITIVE LOGITS

 selected

2.05

 select

1.74

selected

1.69

 ausgewä

1.61

 Selected

1.53

 wybran

1.44

Selected

1.43

 chosen

1.33

 SELECTED

1.29

select

1.26

Activations Density 0.683%

phrases related to research methodologies and search strategies

No Comments

No Known Activations