INDEX

Explanations

technical terms and variables related to data and analysis processes

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

DiCFS-hp and DiCFS-vp

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

AndroidJUnit

-0.84

 purpoſe

-0.81

)"),

-0.79

 pleaſure

-0.77

ollectionView

-0.75

 AssemblyProduct

-0.75

atrième

-0.73

 protoimpl

-0.73

 ſtate

-0.73

Eſ

-0.73

POSITIVE LOGITS

ang

0.40

kin

0.39

AMPTON

0.37

DISCUSSION

0.37

сора

0.36

 بگیر

0.36

 waardoor

0.35

 huesos

0.35

жига

0.35

Activations Density 1.436%

technical terms and variables related to data and analysis processes

DiCFS-hp and DiCFS-vp

No Comments

No Known Activations