INDEX

Explanations

references to data types and memory alignment in coding contexts

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

Gön

-0.61

httphttps

-0.59

AndEndTag

-0.55

 AspNetCore

-0.52

createStatement

-0.51

 defaultstate

-0.50

webElementXpaths

-0.48

 للمعارف

-0.47

 fusca

-0.46

SequentialGroup

-0.46

POSITIVE LOGITS

sizeof

0.67

 sizeof

0.59

ENDIAN

0.36

MessageTagHelper

0.36

AccessFile

0.35

setof

0.35

 mystère

0.34

Layout

0.34

굼

0.33

 soigne

0.33

Activations Density 0.015%

references to data types and memory alignment in coding contexts

No Comments

No Known Activations