INDEX

Explanations

phrases related to understanding and comprehension

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

confirming understanding

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

AnimationsModule

-0.46

outSlope

-0.43

 InputDecoration

-0.41

 OFDb

-0.40

RenderAtEndOf

-0.39

LookAnd

-0.39

MemoryWarning

-0.38

AsUp

-0.36

 informée

-0.36

étude

-0.36

POSITIVE LOGITS

 متعلقه

0.48

 tacit

0.47

튿

0.45

 tarko

0.43

 presup

0.42

Leaving

0.42

 Fleck

0.40

чается

0.40

 misunder

0.40

Essentially

0.39

Activations Density 0.117%

phrases related to understanding and comprehension

confirming understanding

No Comments

No Known Activations