INDEX

Explanations

terms related to human activities and processes, particularly in experimental and investigational contexts

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

during the experiment, battle, or process

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

 GoogleFonts

-0.70

хьтан

-0.60

 ویکی‌پدی

-0.59

 otomatig

-0.56

 iNdEx

-0.55

ьаж

-0.55

acles

-0.54

ackerel

-0.54

 Taktlose

-0.53

("%.

-0.52

POSITIVE LOGITS

 Dabei

0.49

 towarzys

0.43

ModelSerializer

0.42

 deltag

0.37

Dabei

0.37

 Incluye

0.37

 pihaknya

0.36

 défend

0.36

 dabei

0.36

 inoltre

0.36

Activations Density 0.051%

terms related to human activities and processes, particularly in experimental and investigational contexts

during the experiment, battle, or process

No Comments

No Known Activations