INDEX

Explanations

mentions of operators and their actions or roles in a specific process

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

machine operator actions

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

 initComponents

-0.41

----------------

-0.37

ţ

-0.32

 wildflowers

-0.32

GetMapping

-0.31

 encontrados

-0.31

된

-0.30

URING

-0.30

americana

-0.30

ęło

-0.30

POSITIVE LOGITS

 staff

0.60

IsPostBack

0.53

 technician

0.53

 <>",

0.53

staff

0.53

AndEndTag

0.53

RTLI

0.52

 technicians

0.52

 unskilled

0.52

STAFF

0.52

Activations Density 0.066%

mentions of operators and their actions or roles in a specific process

machine operator actions

No Comments

No Known Activations