INDEX

Explanations

code structure and definitions in programming languages

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Sementara

-0.27

 while

-0.24

Meanwhile

-0.22

 between

-0.22

 Erkenntnisse

-0.22

ou

-0.21

cji

-0.21

mi

-0.21

Couldn

-0.20

via

-0.20

POSITIVE LOGITS

RenderAtEndOf

0.85

MemoryWarning

0.79

ніципалі

0.75

ſſung

0.75

<unused16>

0.74

нгред

0.74

<unused74>

0.74

<unused41>

0.74

[@BOS@]

0.73

<unused8>

0.73

Activations Density 0.223%

code structure and definitions in programming languages

No Comments

No Known Activations