INDEX

Explanations

expressions related to assessing and evaluating information or performance

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 désolés

-0.60

tonode

-0.60

SharedCtor

-0.56

 становника

-0.54

roffenen

-0.54

Personensuche

-0.53

 فريبيس

-0.52

BitConverter

-0.52

 Biôgrafia

-0.52

PasswordEncoder

-0.51

POSITIVE LOGITS

 Greg

0.32

Ben

0.31

блю

0.30

Tru

0.30

 promoted

0.30

Modes

0.30

 Marine

0.29

 marine

0.29

Leg

0.29

appré

0.28

Activations Density 0.201%

expressions related to assessing and evaluating information or performance

No Comments

No Known Activations