INDEX

Explanations

specific technical terms and values related to sessions and data attributes

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

genetic mutations and treatments

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

 Join

-0.46

quel

-0.44

major

-0.43

 انج

-0.43

 suga

-0.42

Pol

-0.42

дитесь

-0.42

main

-0.42

difficulty

-0.42

meg

-0.42

POSITIVE LOGITS

adpleegd

0.59

 nahilalakip

0.59

'{@

0.53

 disambiguazione

0.51

 OFDb

0.49

 հղումներ

0.48

 plufieurs

0.47

 دیکھیے

0.47

 actionMode

0.47

ArrowToggle

0.47

Activations Density 0.000%

specific technical terms and values related to sessions and data attributes

genetic mutations and treatments

No Comments

No Known Activations