INDEX

Explanations

questions and phrases related to personal inquiry and preferences

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

questions asking you

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

 suaminya

-0.41

orsese

-0.38

 Վերցված

-0.38

 conmigo

-0.37

 pihaknya

-0.35

tonode

-0.35

 navideña

-0.35

 něj

-0.35

 keluarganya

-0.35

Erreferentziak

-0.34

POSITIVE LOGITS

rungsseite

0.69

 AttributeSet

0.60

 Paglinawan

0.57

sidemargin

0.52

uxxxx

0.49

 passion

0.48

amerikanischer

0.48

 незавершена

0.47

 passionate

0.47

 Alexand

0.46

Activations Density 0.006%

questions and phrases related to personal inquiry and preferences

questions asking you

No Comments

No Known Activations