INDEX

Explanations

technical terms and code-related language in programming and documentation

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

programming keywords

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

 surla

-0.65

 resourceCulture

-0.64

WriteTagHelper

-0.63

posedge

-0.63

 InputDecoration

-0.61

 ProtoMessage

-0.56

SharedCtor

-0.56

setVerticalGroup

-0.56

tranno

-0.55

RemoveField

-0.55

POSITIVE LOGITS

 OFDb

0.36

eseorang

0.32

 Cientí

0.28

 Ambiental

0.26

 going

0.26

 forças

0.25

 intérieure

0.24

öz

0.24

 chiuso

0.24

 zprávy

0.23

Activations Density 0.373%

technical terms and code-related language in programming and documentation

programming keywords

No Comments

No Known Activations