INDEX

Explanations

phrases indicating movement or transition between locations or contexts

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

prepositions followed by a noun

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

endif

-0.35

ğan

-0.27

 podio

-0.27

来自

-0.26

ENDIF

-0.26

joy

-0.26

open

-0.26

 classic

-0.26

 joys

-0.25

the

-0.25

POSITIVE LOGITS

yntaxException

0.81

URLException

0.78

 verſch

0.69

StoryboardSegue

0.69

 referrerpolicy

0.68

 الحره

0.67

openConnection

0.67

AutoModerator

0.66

󠁴

0.66

andExpect

0.65

Activations Density 0.054%

phrases indicating movement or transition between locations or contexts

prepositions followed by a noun

No Comments

No Known Activations