INDEX

Explanations

code segments related to state management and variable updates

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

while and count

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

-0.60

 estekak

-0.59

 kasarigan

-0.49

RRect

-0.48

脚注の使い方

-0.48

таратура

-0.45

 uska

-0.44

ParallelGroup

-0.42

 ligiloj

-0.42

 muñeco

-0.41

POSITIVE LOGITS

EditorBrowsable

0.44

isStatic

0.38

 endwhile

0.37

sise

0.36

 Instability

0.35

 prosi

0.35

wise

0.35

hú

0.35

hoea

0.34

bewerken

0.34

Activations Density 0.018%

code segments related to state management and variable updates

while and count

No Comments

No Known Activations