INDEX

Explanations

specific numeric patterns or identifiers related to datasets and entities

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

numbers, codes, or identifiers

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

CPtr

-0.59

ConstraintMaker

-0.56

 <=",

-0.54

ngdoc

-0.53

 Rivas

-0.50

 Goode

-0.50

dafx

-0.48

 hotline

-0.46

astify

-0.46

 lenker

-0.45

POSITIVE LOGITS

 FontWeight

0.37

IllegalArgument

0.34



0.33

 IllegalArgument

0.31

 intptr

0.30

 bcrypt

0.29

nvarchar

0.29

0.28

RouterModule

0.28

Activations Density 0.188%

specific numeric patterns or identifiers related to datasets and entities

numbers, codes, or identifiers

No Comments

No Known Activations