INDEX

Explanations

code-related keywords and constructs

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

Bitwise shift and mask operations

np_acts-logits-general · gemini-2.0-flash

C++ preprocessor directives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.28.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

windowFixed

-0.74

 queſta

-0.73

 '\\;'

-0.73

 ProtoMessage

-0.71

rungsseite

-0.69

extAlignment

-0.68

ArrowToggle

-0.68

ItemBackground

-0.68

IUrlHelper

-0.67

 EconPapers

-0.64

POSITIVE LOGITS

BeforeMethod

0.35

="{{$

0.33

 mentor

0.32

 decedent

0.31

est

0.30

 astore

0.30

he

0.29

land

0.28

hen

0.28

Activations Density 0.377%

code-related keywords and constructs

Bitwise shift and mask operations

C++ preprocessor directives

No Comments

No Known Activations