INDEX

Explanations

programming-related terms, particularly those associated with options and parameters in function calls

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

foreign languages and code syntax

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

RegressionTest

-0.69

الحياه

-0.67

ReusableCell

-0.64

 CanadaChoose

-0.63

 MainAxisSize

-0.63

MLLoader

-0.59

ArgsConstructor

-0.59

 ſeinen

-0.57

 ſeine

-0.56

évaluateur

-0.56

POSITIVE LOGITS

Ārējās

0.33

 including

0.31

 except

0.30

orElseThrow

0.30

来歴

0.30

 kecuali

0.30

förs

0.29

tvguidetime

0.29

 sauf

0.28

documentclass

0.28

Activations Density 0.007%

programming-related terms, particularly those associated with options and parameters in function calls

foreign languages and code syntax

No Comments

No Known Activations