INDEX

Explanations

high numerical values likely related to measurements or identifiers in a dataset

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Новости

-0.45

InView

-0.44

 FILTER

-0.43

 Circ

-0.43

 reiz

-0.42

 Daley

-0.42

 blanco

-0.41

 renewal

-0.41

ự

-0.41

 Blanca

-0.40

POSITIVE LOGITS

IndentedString

1.06

windowFixed

1.01

RegressionTest

0.92

GEBURTSDATUM

0.84

 AssemblyCulture

0.83

 eighth

0.80

ThroughAttribute

0.74

0.73

 Eighth

0.72

⑧

0.71

Activations Density 0.185%

high numerical values likely related to measurements or identifiers in a dataset

No Comments

No Known Activations