INDEX

Explanations

key phrases indicating quantitative significance or frequency

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

emphasizing or particularizing

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

 فريبيس

-0.60

OCCURRED

-0.59

 initComponents

-0.59

nosaurus

-0.57

 nahilalakip

-0.56

 CreateTagHelper

-0.55

SPATH

-0.54

'{@

-0.52

 onAnimation

-0.52

AsUp

-0.52

POSITIVE LOGITS

 especialmente

0.45

 especial

0.41

 szczególnie

0.39

 especially

0.36

 zwłaszcza

0.35

GEBURTSDATUM

0.34

czegó

0.34

特に

0.34

 zejména

0.33

 notamment

0.33

Activations Density 0.095%

key phrases indicating quantitative significance or frequency

emphasizing or particularizing

No Comments

No Known Activations