INDEX

Explanations

references to changes in statistical or numerical data, particularly regarding increases or trends

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

conditions and states

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

AnchorTagHelper

-0.44

egt

-0.42

Angeb

-0.41

 anuales

-0.40

 Mérimée

-0.39

게

-0.38

 dogmatic

-0.38

impresa

-0.38

bindingNavigator

-0.38

 Michael

-0.38

POSITIVE LOGITS

ſelf

0.54

neſs

0.49

 trend

0.47

fromnode

0.46

IntoConstraints

0.45

 infection

0.44

 محفوظة

0.44

tvguidetime

0.44

 dieß

0.43

 itſelf

0.43

Activations Density 0.007%

references to changes in statistical or numerical data, particularly regarding increases or trends

conditions and states

No Comments

No Known Activations