INDEX

Explanations

significant findings or results from scientific studies

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

data and results suggesting findings

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

WarningLevel

-0.60

 brancas

-0.57

 pioggia

-0.53

pulseira

-0.53

 quelcon

-0.53

NetworkInfo

-0.52

참고

-0.51

 acoper

-0.50

 colazione

-0.50

errHandler

-0.50

POSITIVE LOGITS

 تانيه

0.40

 demostr

0.36

证

0.36

 performed

0.34

 confirme

0.33

transQ

0.33

 évidence

0.32

 phenotypes

0.32

 provas

0.31

initComponents

0.31

Activations Density 0.682%

significant findings or results from scientific studies

data and results suggesting findings

No Comments

No Known Activations