INDEX

Explanations

references to specific entities, queries, or cases within a legal or scientific context

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

specific issues and 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 @ 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

 obrazu

-0.27

 laufen

-0.26

 conjunta

-0.24

 dotato

-0.24

Gr

-0.24

and

-0.24

 konzent

-0.24

 haline

-0.24

 understanding

-0.24

now

-0.24

POSITIVE LOGITS

GEBURTSDATUM

0.76

 surla

0.74

 ब्रेकडाउन

0.68

 мәкал

0.65

liceerd

0.63

tagHelperRunner

0.62

 typelib

0.60

 MainAxisSize

0.59

IUrlHelper

0.58

majánló

0.57

Activations Density 0.073%

references to specific entities, queries, or cases within a legal or scientific context

specific issues and technical terms

No Comments

No Known Activations