INDEX

Explanations

punctuation marks and formatting symbols

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

code formatting or variables

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

 للاسماء

-0.74

nestjs

-0.64

 يتيمه

-0.63

SharedCtor

-0.60

Tikang

-0.60

ScopeManager

-0.57

ніципалі

-0.57

ыгана

-0.53

aarrggbb

-0.52

 HasFactory

-0.49

POSITIVE LOGITS

 pysty

0.31

 julk

0.30

 merak

0.30

 näky

0.29

trek

0.29

 Retired

0.29

\{\\

0.29

prin

0.29

ünst

0.29

 jakie

0.29

Activations Density 0.064%

punctuation marks and formatting symbols

code formatting or variables

No Comments

No Known Activations