INDEX

Explanations

instances of high activation values, likely indicating important features or commands in the code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

↵↵

-1.15

[toxicity=0]

-0.79

-0.65

-0.63

ed

-0.63

-0.62

awt

-0.62

↵

-0.61

ing

-0.58

POSITIVE LOGITS

rungsseite

1.04

verwijspagina

0.97

 iſt

0.94

 AssemblyVersion

0.92

]-->

0.88

Portály

0.88

 gogh

0.88

 ―――――

0.86

にほんブログ村

0.85

 FetchType

0.85

Activations Density 0.019%

instances of high activation values, likely indicating important features or commands in the code

No Comments

No Known Activations