INDEX

Explanations

terms related to mathematical or logical structure and organization

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

extAlignment

-0.52

 nakalista

-0.51

новниш

-0.51

 avoient

-0.50

complexContent

-0.50

 pleaſure

-0.49

 bezeichneter

-0.48

 DriverManager

-0.48

:✨

-0.47

 étoient

-0.46

POSITIVE LOGITS

utor

0.39

Microkernel

0.36

CharacterOffset

0.35

tew

0.35

zeuge

0.34

rrggbb

0.33

 ffilmiau

0.33

TagHelper

0.33

 otomatig

0.33

../../../

0.32

Activations Density 0.054%

terms related to mathematical or logical structure and organization

No Comments

No Known Activations