INDEX

Explanations

elements related to memory addresses and their operations in programming

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

memory address

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

onOptions

-0.40

 Anſ

-0.39

 cast

-0.37

 occaf

-0.36

 endforeach

-0.35

 reaſon

-0.34

 Cast

-0.34

 faſt

-0.34

دانشنامهٔ

-0.34

 fhort

-0.34

POSITIVE LOGITS

ConstraintMaker

0.60

 disambiguazione

0.59

 AssemblyVersion

0.56

Assured

0.53

setof

0.50

Koordinaten

0.49

sidemargin

0.49

 acceder

0.48

 accessing

0.48

 fiume

0.48

Activations Density 0.439%

elements related to memory addresses and their operations in programming

memory address

No Comments

No Known Activations