INDEX

Explanations

expressions that convey strong emotional impact or significant reactions

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

strong positive descriptions

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 tuot

-0.41

 تضيفلها

-0.41

 Municipalidad

-0.35

 hipó

-0.35

 lapangan

-0.34

 vuotta

-0.34

 nahilalakip

-0.34

 chercheurs

-0.34

 pimpinan

-0.33

 précédente

-0.32

POSITIVE LOGITS

ChildScrollView

0.81

 مشين

0.73

 restTemplate

0.55

tanleria

0.54

 الرياضيه

0.54

ocarc

0.54

ngrx

0.54

拗

0.53

 ujednoznacz

0.52

TestBed

0.52

Activations Density 0.151%

expressions that convey strong emotional impact or significant reactions

strong positive descriptions

No Known Activations