INDEX

Explanations

structured data representations or tabular formats

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

-0.52

//

-0.50

\\

-0.43

↵

-0.42

<strong>

-0.42

↵↵↵

-0.42

/*

-0.41

POSITIVE LOGITS

ValueStyle

0.93

WriteTagHelper

0.86

writeFieldEnd

0.77

SourceChecksum

0.73

 resourceCulture

0.71

reportWebVitals

0.69

ConstraintMaker

0.69

 مرئيه

0.68

>\<^

0.67

 Савезне

0.67

Activations Density 1.623%

structured data representations or tabular formats

No Known Activations