INDEX

Explanations

references to code and programming, particularly modifications and code functionality

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

technical or programming code

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

elemField

-0.56

afficheront

-0.44

AISSEE

-0.36

 işlem

-0.34

CppMethod

-0.30

はじめに

-0.29

"]]

-0.27

mbggenerated

-0.27

oredCriteria

-0.26

製

-0.26

POSITIVE LOGITS

 resourceCulture

0.54

LLocation

0.53

Cyfeiriadau

0.52

ativement

0.52

 InputDecoration

0.50

┓

0.50

 lenker

0.50

TagMode

0.49

Buc

0.48

UnknownFieldSet

0.48

Activations Density 0.228%

references to code and programming, particularly modifications and code functionality

technical or programming code

No Comments

No Known Activations