INDEX

Explanations

keywords and phrases related to transcripts and language usage

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

proper names 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

Datuak

-0.47

 незавершена

-0.46

UnusedPrivate

-0.45

 وتسجيلات

-0.41

RTDA

-0.41

lapia

-0.40

HasAnnotation

-0.40

sizeCache

-0.39

Tikang

-0.38

 incapac

-0.38

POSITIVE LOGITS

hyrchwyd

0.44

calça

0.43

 Signalez

0.43

antMatchers

0.42

 vôtre

0.41

 tetto

0.41

ServiceTest

0.40

 Italijani

0.40

rawDesc

0.40

laß

0.40

Activations Density 0.302%

keywords and phrases related to transcripts and language usage

proper names and technical terms

No Comments

No Known Activations