INDEX

Explanations

terms related to technical specifications and frameworks in engineering and legal contexts

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

nouns ending clauses

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

 ویکی‌پدی

-0.58

ьаж

-0.56

NameInMap

-0.53

RTGC

-0.51

Chham

-0.49

RTSC

-0.47

RegistryLite

-0.47

 pinulongan

-0.45

 ivelany

-0.43

testify

-0.43

POSITIVE LOGITS

Carriera

0.48

 InputDecoration

0.45

ReusableCell

0.42

 bezeichneter

0.42

SourceChecksum

0.40

 ujednoznacz

0.39

 EconPapers

0.39

InputBorder

0.35

verwijspagina

0.35



0.35

Activations Density 0.082%

terms related to technical specifications and frameworks in engineering and legal contexts

nouns ending clauses

No Comments

No Known Activations