INDEX

Explanations

non-standard formatting and technical terms related to programming or legal documentation

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

legal and scientific processes

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

 Bono

-0.36

OCCURRED

-0.35

 indoss

-0.32

authier

-0.32

langsung

-0.31

 EconPapers

-0.31

 voorbij

-0.31

 invloed

-0.31

 elliptic

-0.31

 constrain

-0.31

POSITIVE LOGITS

 ब्रेकडाउन

0.57

 transférez

0.57

 vPvB

0.57

awtextra

0.56

 تضيفلها

0.56

 BoxFit

0.54

sidemargin

0.54

gameserver

0.53

arschu

0.53

 continúas

0.53

Activations Density 0.047%

non-standard formatting and technical terms related to programming or legal documentation

legal and scientific processes

No Comments

No Known Activations