INDEX

Explanations

specific keywords or terminologies commonly associated with scientific or technical content

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

foreign words and technical terms

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

 nakalista

-0.58

 незавершена

-0.57

KURZBESCHREIBUNG

-0.54

 مشين

-0.54

Personendaten

-0.54

DotNetBar

-0.52

intios

-0.51

 consultato

-0.50

expandindo

-0.50

dini

-0.50

POSITIVE LOGITS

 חיצוניים

0.40

const

0.35



0.34

 defaultstate

0.33

osoba

0.32

 hjär

0.30

ré

0.30

less

0.30

 terrestre

0.30

 apag

0.30

Activations Density 0.036%

specific keywords or terminologies commonly associated with scientific or technical content

foreign words and technical terms

No Comments

No Known Activations