INDEX

Explanations

quantitative measurements and numerical data

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

numbers and units

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

 rencontres

-0.40

 następnie

-0.40

 événements

-0.39

 chaus

-0.39

 célèbres

-0.39

Sinon

-0.38

 "..\..\..\

-0.38

 bienfaits

-0.38

 apud

-0.38

 administratif

-0.38

POSITIVE LOGITS

+#+#

0.51

 lousy

0.47

 للمعارف

0.46

ReusableCell

0.46

hdashline

0.46

 BytesLike

0.45

Erreferentziak

0.45

IUrlHelper

0.44

 nicer

0.44

 decent

0.43

Activations Density 0.037%

quantitative measurements and numerical data

numbers and units

No Comments

No Known Activations