INDEX

Explanations

comments and documentation indicators in code snippets

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

code comments

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

Datuak

-0.57

ftagPool

-0.52

Portale

-0.45

stylers

-0.44

 crossorigin

-0.44

TRIBUN

-0.44

انجليز

-0.43

TemporalType

-0.42

 pinulongan

-0.42

 arşivlendi

-0.42

POSITIVE LOGITS

 initComponents

0.53

 GenerationType

0.50

GIVEREF

0.47

Opere

0.43

 Catania

0.43

énario

0.41

 söyledi

0.40

PYX

0.39

spesies

0.38

Assemblée

0.38

Activations Density 0.029%

comments and documentation indicators in code snippets

code comments

No Comments

No Known Activations