INDEX

Explanations

references to experimental results and statistical data in scientific contexts

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

numbers and symbols

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

UserScript

-0.81

bootstrapcdn

-0.61

OGND

-0.60

WireFormatLite

-0.60

aarrggbb

-0.60

Transcripción

-0.59

Transkript

-0.59

Gön

-0.59

RTEE

-0.58

exitRule

-0.57

POSITIVE LOGITS

 dokład

0.39

 Stimme

0.38

 ENGINEER

0.37

 recientes

0.37

 Proben

0.37

 légèrement

0.36

 ciąży

0.36

 gewissen

0.36

 Typen

0.36

 adicionales

0.35

Activations Density 0.002%

references to experimental results and statistical data in scientific contexts

numbers and symbols

No Comments

No Known Activations