INDEX

Explanations

technical terms and processes related to product functionality and usability

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

simple instructions, simple 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

localctx

-0.52

 uLocal

-0.50

AddTagHelper

-0.49

ucco

-0.46

PickerController

-0.46

createConnection

-0.45

StoryboardSegue

-0.45

-0.44

 estekak

-0.42

 {}".

-0.42

POSITIVE LOGITS

 Chwiliwch

0.46

always

0.41

 metálico

0.39

 wystarczy

0.39

 zawsze

0.38

 always

0.38

melidir

0.38

ďaka

0.38

 trasparente

0.37

cée

0.37

Activations Density 0.012%

technical terms and processes related to product functionality and usability

simple instructions, simple actions

No Comments

No Known Activations