INDEX

Explanations

patterns of whitespace and other delimiters in text

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

user-specified environment

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

Tikang

-0.74

ViewFeatures

-0.69

 ब्रेकडाउन

-0.68

Personensuche

-0.65

abestanden

-0.63

tvguidetime

-0.62

 resourceCulture

-0.61

 AssemblyTitle

-0.58

 jsPsych

-0.57

 Walkover

-0.56

POSITIVE LOGITS

ModelState

0.34

ípio

0.34

 StatelessWidget

0.34

 Touren

0.33

nextState

0.31

cascade

0.30

мка

0.30

Kreis

0.29

Something

0.29

 chấm

0.29

Activations Density 0.012%

patterns of whitespace and other delimiters in text

user-specified environment

No Comments

No Known Activations