INDEX

Explanations

technical terms and class definitions in code or programming contexts

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

code identifiers

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

 MaterialApp

-0.39

 then

-0.38

ouille

-0.35

then

-0.32

 beast

-0.32

JvmStatic

-0.32

ga

-0.31

ad

-0.31

Ins

-0.31

Multi

-0.31

POSITIVE LOGITS

adaptiveStyles

0.66

 EconPapers

0.59

 betweenstory

0.57

posedge

0.57

rungsseite

0.56

OGND

0.54

styleType

0.53

verwijspagina

0.52

GEBURTSDATUM

0.52

цездатний

0.52

Activations Density 0.002%

technical terms and class definitions in code or programming contexts

code identifiers

No Comments

No Known Activations