INDEX

Explanations

past actions and plans related to a project or process

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

originally or supposed to

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

MessageOf

-0.57

 esternos

-0.52

 Autorisations

-0.50

ThroughAttribute

-0.46

msgTypes

-0.45

tanleria

-0.45

RenderAtEndOf

-0.45

swig

-0.43

исленность

-0.42

 Италијани

-0.42

POSITIVE LOGITS

WithIOException

0.50

findpost

0.49

 Grüsse

0.49

+#+#

0.43

 Betracht

0.40

 słucha

0.39

zeitige

0.38

 bowiem

0.38

 encu

0.38

 wcześ

0.36

Activations Density 0.041%

past actions and plans related to a project or process

originally or supposed to

No Comments

No Known Activations