INDEX

Explanations

instances of high-stakes legal terms or scenarios

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

technical terms

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

MathML

-0.63

الحياه

-0.56

 Chwiliwch

-0.53

 Meal

-0.51

anken

-0.51

 Share

-0.49

-------------</

-0.49

wage

-0.48

 <=",

-0.48

 Sort

-0.48

POSITIVE LOGITS

 ویکی‌پدیای

0.37

 nakalista

0.37

\{\\

0.36

 cuarzo

0.34

Демографія

0.33

PROTOBUF

0.32

 définiti

0.32

 AssemblyTitle

0.32

GeneratedValue

0.32

 Kalifor

0.31

Activations Density 0.366%

instances of high-stakes legal terms or scenarios

technical terms

No Comments

No Known Activations