INDEX

Explanations

code-related elements, specifically focusing on method definitions and control flow structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

racene

-0.53

ankin

-0.51

inkin

-0.50

AMAZ

-0.47

 sinar

-0.47

 raya

-0.47

mê

-0.47

mito

-0.47

 peine

-0.46

teries

-0.45

POSITIVE LOGITS

public

1.18

 public

1.02

 Public

0.83

 PUBLIC

0.83

 pubblici

0.83

WithIOException

0.80

 Pública

0.80

publ

0.78

 تضيفلها

0.77

Public

0.76

Activations Density 0.027%

code-related elements, specifically focusing on method definitions and control flow structures

No Comments

No Known Activations