INDEX

Explanations

constructs involving conditional logic and functions in programming code

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

code and non-english words

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



-0.63

StructEnd

-0.60

RegressionTest

-0.60

SBATCH

-0.59

tagHelperRunner

-0.59

 AssemblyCulture

-0.59

 AssemblyCompany

-0.58

-0.56

 estekak

-0.55

RectangleBorder

-0.54

POSITIVE LOGITS

OGND

0.47

mauri

0.45

الدراسه

0.44

 Neub

0.43

形で

0.41

 inderdaad

0.41

makeConstraints

0.41

 ואת

0.41

じて

0.40

RTCK

0.40

Activations Density 0.011%

constructs involving conditional logic and functions in programming code

code and non-english words

No Comments

No Known Activations