INDEX

Explanations

programming and technical terms related to execution and evaluation processes

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

amendment is applied retroactively

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

GEBURTSDATUM

-0.54

awtextra

-0.49

 pleaſure

-0.43

 AssemblyTitle

-0.42

 purpoſe

-0.42

 Exacts

-0.41

➯

-0.39

ualaikum

-0.38

 DBNull

-0.38

 houſe

-0.36

POSITIVE LOGITS

OGND

0.57

 Chwiliwch

0.52

 незавершена

0.48

uxxxx

0.47

ticons

0.45

Mob

0.41

principalTable

0.40

mige

0.39

IALES

0.39

 مشين

0.39

Activations Density 0.112%

programming and technical terms related to execution and evaluation processes

amendment is applied retroactively

No Comments

No Known Activations