INDEX

Explanations

phrases related to quantitative increases or changes

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

BeginContext

-0.64

 indisponible

-0.52

 Wund

-0.47

subsubsection

-0.46

Atsauces

-0.45

SequentialGroup

-0.44

leroi

-0.44

onoi

-0.44

isempty

-0.44

льтра

-0.43

POSITIVE LOGITS

 increase

2.97

 increased

2.74

increase

2.63

 Increase

2.57

Increase

2.51

 increases

2.49

increased

2.44

 Increased

2.33

 INCREASE

2.32

Increased

2.24

Activations Density 1.761%

phrases related to quantitative increases or changes

No Comments

No Known Activations