INDEX

Explanations

specific programming constructs and syntax elements related to categorization and data processing

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

Code, variables, or mathematical expressions

np_acts-logits-general · gemini-2.0-flash

$ characters and punctuation

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 faſt

-0.59

Efq

-0.56

featureID

-0.56

ConstraintMaker

-0.52

 يتيمه

-0.52

 ſta

-0.50

 oprot

-0.50

 oock

-0.50

 cref

-0.50

 SWIG

-0.47

POSITIVE LOGITS

ględ

0.37

jenost

0.35

sim

0.35

CodedInputStream

0.35

findpost

0.35

 whereas

0.35

 defaultstate

0.34

tritt

0.34

ENAS

0.33

 przy

0.33

Activations Density 1.137%

specific programming constructs and syntax elements related to categorization and data processing

Code, variables, or mathematical expressions

$ characters and punctuation

No Comments

No Known Activations