INDEX

Explanations

actions and processes related to functionality and outcomes

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

verb-object relationships

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

Rüyada

-0.41

Ɓ

-0.41

ImageContext

-0.40

ceğine

-0.40

upol

-0.39

saraba

-0.38

 Future

-0.38

TestRunner

-0.38

تقاوى

-0.37

########.

-0.36

POSITIVE LOGITS

complexContent

0.57

InputBorder

0.48

 ब्रेकडाउन

0.46

Jeografia

0.44

 nakalista

0.44

 indisponible

0.42

gridx

0.41

cartão

0.41

StrictEqual

0.41

neté

0.41

Activations Density 0.096%

actions and processes related to functionality and outcomes

verb-object relationships

No Comments

No Known Activations