INDEX

Explanations

references to minor details and moments that contribute to larger experiences or narratives

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

small things

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

 transférez

-0.57

EFIAPI

-0.54

coque

-0.46

clerView

-0.46

лтемелер

-0.46

厲害

-0.46

GOTREF

-0.45

ffeine

-0.44

цезда

-0.43

⟬

-0.43

POSITIVE LOGITS

 små

0.74

 piccole

0.70

 small

0.68

Small

0.68

 pequeños

0.67

 pequenos

0.66

 tiny

0.66

 piccoli

0.65

 pequeñas

0.65

small

0.65

Activations Density 0.167%

references to minor details and moments that contribute to larger experiences or narratives

small things

No Known Activations