INDEX

Explanations

statistical terms and measurements related to data analysis

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

normalization and calibration

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

AndroidJUnit

-0.53

iprot

-0.51

ArrowToggle

-0.49

alakip

-0.45

SPJ

-0.44

internalType

-0.44

endpush

-0.43

SequentialGroup

-0.43

msgSender

-0.43

ThroughAttribute

-0.43

POSITIVE LOGITS

GEBURTSDATUM

0.50

 alér

0.34

 Sheeran

0.33

 douceur

0.32

 leeren

0.31

 edelstahl

0.31

 iniziale

0.31

 mínimos

0.31

 Tasche

0.31

 punctual

0.31

Activations Density 0.115%

statistical terms and measurements related to data analysis

normalization and calibration

No Comments

No Known Activations