INDEX

Explanations

phrases indicating superiority or high quality

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

best choice/option/alternative

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

 المعيارى

-0.55

 GenerationType

-0.50

IntOverflow

-0.49

🖨

-0.49

Brainz

-0.47

especie

-0.46

exh

-0.46

rungsseite

-0.46

grained

-0.45

 Himo

-0.43

POSITIVE LOGITS

 iprot

0.38

migrationBuilder

0.38

posób

0.36

AddTagHelper

0.34

StoreMessageInfo

0.34

イヤル

0.34

Kaynakça

0.32

LookAnd

0.31

 GlobalKey

0.31

InjectAttribute

0.31

Activations Density 0.004%

phrases indicating superiority or high quality

best choice/option/alternative

No Comments

No Known Activations