INDEX

Explanations

keywords indicating bad programming practices and errors in code

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

code, errors, and conversation

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

webElementXpaths

-1.42

 propOrder

-1.34

HasAnnotation

-0.60

LLocation

-0.55

 uLocal

-0.40

 rzeczy

-0.30

 estekak

-0.30

 erlaubt

-0.30

 Sekunden

-0.29

-0.28

POSITIVE LOGITS

OGND

3.77

GEBURTSDATUM

3.59

 виправивши

3.05

 ostavi

1.76

abestanden

1.69

 дописавши

1.60

 bezeichneter

1.55

 HttpNotFound

1.33

tagHelperRunner

1.28

expandindo

1.28

Activations Density 0.000%

keywords indicating bad programming practices and errors in code

code, errors, and conversation

No Comments

No Known Activations