INDEX

Explanations

references to significant comparisons and statistics

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

suggesting or indicating outcomes

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

 betweenstory

-0.48

WithIOException

-0.47

GEBURTSDATUM

-0.46

AsUp

-0.45

ôtel

-0.43

 chanter

-0.43

 nhàng

-0.43

setCellStyle

-0.41

ItemBackground

-0.41

 createState

-0.41

POSITIVE LOGITS

 AssemblyCulture

0.46

 signals

0.42

 juger

0.42

 INDIC

0.42

 indicates

0.42

 indication

0.39

 zust

0.38

 indiquer

0.38

 indicador

0.37

stddef

0.37

Activations Density 0.058%

references to significant comparisons and statistics

suggesting or indicating outcomes

No Comments

No Known Activations