INDEX

Explanations

Java programming language related terms and imports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

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

 grades

-0.40

 every

-0.36

ash

-0.35

 enim

-0.35

 grade

-0.35

 preload

-0.35

 saja

-0.35

by

-0.34

之内

-0.34

 everyone

-0.33

POSITIVE LOGITS

Datuak

0.85

rungsseite

0.81

MessageTagHelper

0.77

 disambiguazione

0.75

 المعيارى

0.74

 Administrativna

0.74

 विश्वसनीयता

0.71

0.70

 censiti

0.69



0.69

Activations Density 0.005%

Java programming language related terms and imports

No Comments

No Known Activations