INDEX

Explanations

conditional expressions and comparisons within code

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

user input start

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

ThroughAttribute

-0.57

 artem

-0.57

μβρίου

-0.55

 TextInputType

-0.53

Senti

-0.53

URLException

-0.53

 propOrder

-0.52

rdom

-0.52

 Sancti

-0.51

 condens

-0.51

POSITIVE LOGITS

setVerticalGroup

0.45

PreferredItem

0.44

 bahía

0.42

 Exactos

0.39

sidemargin

0.39

 estrecha

0.37

 étro

0.37

 étroite

0.37

 isolado

0.34

npos

0.34

Activations Density 0.005%

conditional expressions and comparisons within code

user input start

No Comments

No Known Activations