INDEX

Explanations

references to navigating and summarizing information in written content

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

Code snippets or programming-related text

np_acts-logits-general · gemini-2.0-flash

skip, extract, sampling, headers, assumed

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 prefeito

-0.36

 mourut

-0.33

 fondateur

-0.32

 pédagogique

-0.31

issait

-0.31

 courriel

-0.31

 Einfluß

-0.30

 nôtre

-0.30

 pulseras

-0.29

 Hauteur

-0.29

POSITIVE LOGITS

CppMethod

0.77

 AssemblyCulture

0.71

writeFieldEnd

0.65

 وتسجيلات

0.63

 skip

0.60

SourceChecksum

0.60

 skipped

0.60

skip

0.57

AsUp

0.57

 summaries

0.56

Activations Density 0.379%

references to navigating and summarizing information in written content

Code snippets or programming-related text

skip, extract, sampling, headers, assumed

No Comments

No Known Activations