INDEX

Explanations

elements related to coding structures or syntax

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

calculus and 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

mergeFrom

-0.63

ArrowToggle

-0.61

InjectAttribute

-0.58

fjspx

-0.57

uxxxx

-0.56

oredCriteria

-0.53

argon

-0.50

 للاسماء

-0.49

Autoritní

-0.49

astify

-0.48

POSITIVE LOGITS

awtextra

0.41

 Dafür

0.37

addPreferredGap

0.34

katapos

0.32

complexContent

0.32

 defaultstate

0.31

ösungen

0.31

>{@

0.31

ampingi

0.31

 inderdaad

0.31

Activations Density 0.001%

elements related to coding structures or syntax

calculus and code

No Comments

No Known Activations