INDEX

Explanations

patterns and structures related to tabular or structured data formatting

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

column names and data

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

LookAnd

-0.69

pulseira

-0.62

calça

-0.61

 BorderSide

-0.61

ChildScrollView

-0.60

 BoxFit

-0.58

 AssemblyCulture

-0.57

bootstrapcdn

-0.57

 biß

-0.56

Clik

-0.56

POSITIVE LOGITS

CppMethod

0.30

ana

0.30

zac

0.28

mó

0.28

 components

0.27

cab

0.27

 Heimat

0.27

hnia

0.27

gi

0.26

 intit

0.26

Activations Density 0.019%

patterns and structures related to tabular or structured data formatting

column names and data

No Comments

No Known Activations