INDEX

Explanations

programming constructs related to data structure manipulation and mapping operations

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

map ordering

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

setVerticalGroup

-0.73

IntoConstraints

-0.65

 snippetHide

-0.63

:✨

-0.62

FunctionFlags

-0.61

AndEndTag

-0.56

 initComponents

-0.55

MessageOf

-0.54

fjspx

-0.54

'{@

-0.54

POSITIVE LOGITS

 Poids

0.36

ForValue

0.34

学院

0.32

 aprende

0.31

ReturnValue

0.30

 chambres

0.29

 Lerner

0.29

 conjug

0.28

 associates

0.28

Expiration

0.28

Activations Density 0.020%

programming constructs related to data structure manipulation and mapping operations

map ordering

No Comments

No Known Activations