INDEX

Explanations

references to identical or similar entities in various contexts

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

same or single item

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

BeforeEach

-0.38

 stuff

-0.36

Népesség

-0.33

 רבה

-0.32

անում

-0.32

 Main

-0.32

 ferons

-0.32

webElementXpaths

-0.32

How

-0.30

 About

-0.30

POSITIVE LOGITS

oredCriteria

0.63

 snippetHide

0.53

 createSprite

0.52

onViewCreated

0.50

 ſind

0.50

Same

0.50

 ujednoznacz

0.49

 Վերցված

0.48

SequentialGroup

0.48

SAME

0.48

Activations Density 0.044%

references to identical or similar entities in various contexts

same or single item

No Comments

No Known Activations