INDEX

Explanations

structural elements in code, particularly block and function definitions

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

code structures

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

 EconPapers

-0.60

HtmlAttribute

-0.53

CloseOperation

-0.50

Gweler

-0.50

 Riproduzione

-0.48

addGap

-0.48

 appelez

-0.48

躇

-0.47

tonode

-0.47



-0.47

POSITIVE LOGITS

 חיצוניים

0.50

awtextra

0.47

protoimpl

0.46

 campeones

0.43

StructEnd

0.40

 disponibilités

0.40

;->

0.40

 EventArgs

0.40

 quedando

0.40

 autorytatywna

0.39

Activations Density 0.009%

structural elements in code, particularly block and function definitions

code structures

No Comments

No Known Activations