INDEX

Explanations

references to structured data lists and elements within a programming context

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

mkern or spacing tokens

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 подвод

-0.42

MessageOf

-0.42

esistenza

-0.40

esercito

-0.40

UrlResolution

-0.39

PositiveButton

-0.38

ToAction

-0.37

Suerte

-0.36

 ferroviaire

-0.36

tellung

-0.36

POSITIVE LOGITS

ValueStyle

0.69

 autorytatywna

0.59

存于互联网档案馆

0.55

 nakalista

0.41

CodeDom

0.40

حوالہ

0.40

complexContent

0.39

naden

0.38

 intptr

0.38

RegressionTest

0.37

Activations Density 0.007%

references to structured data lists and elements within a programming context

mkern or spacing tokens

No Comments

No Known Activations