INDEX

Explanations

elements related to structured data and coding syntax

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

archaic spellings

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

+#+

-0.66

+#+#

-0.61

 Excess

-0.49

oplasmic

-0.47

tonsoft

-0.47

olverine

-0.47

lasher

-0.47

bono

-0.46

 Laser

-0.46

########.

-0.46

POSITIVE LOGITS

 houſe

0.44

WithMany

0.41

 disambiguazione

0.40

ſelves

0.39

 Grecs

0.35

ſelf

0.35

 ſte

0.34

 Conſ

0.34

دانشنامهٔ

0.34

 Houſe

0.34

Activations Density 0.031%

elements related to structured data and coding syntax

archaic spellings

No Comments

No Known Activations