INDEX

Explanations

characters and symbols used in programming or mathematical expressions

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

NameInMap

-0.65

Życiorys

-0.60

complexContent

-0.59

 незавершена

-0.58

 gynhyrchwyd

-0.57

InteropServices

-0.55

 Autorisations

-0.54

ViewFeatures

-0.54

SaveChangesAsync

-0.52

ſchaft

-0.50

POSITIVE LOGITS

setVerticalGroup

0.46

 Sobald

0.35

 Überg

0.34

 Öffnen

0.33

bnf

0.32

 hjäl

0.32

样

0.31

 sebanyak

0.31

 expédi

0.30

 droom

0.30

Activations Density 0.041%

characters and symbols used in programming or mathematical expressions

No Comments

No Known Activations