INDEX

Explanations

programming-related terms and function calls

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

ValueStyle

-0.61

ьаж

-0.55

FailureListener

-0.54

youts

-0.52

politan

-0.52

DoubleQuotes

-0.50

 reflectors

-0.49

 étoit

-0.48

 Administrativna

-0.48

acidade

-0.47

POSITIVE LOGITS

contentLoaded

0.43

 informée

0.40

prepareStatement

0.39

URLConnection

0.36

 mence

0.34

 Normdatei

0.33

consulté

0.32

таратура

0.31

cstdio

0.30

تقاوى

0.30

Activations Density 0.004%

programming-related terms and function calls

No Comments

No Known Activations