INDEX

Explanations

references to numerical values associated with various attributes or characteristics

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

 Wetter

-0.44

Chham

-0.43

 Ernst

-0.41

 Tarjeta

-0.39

 Santos

-0.39

mbols

-0.39

lig

-0.39

ﻰ

-0.38

wetter

-0.38

Jîn

-0.38

POSITIVE LOGITS

setVerticalGroup

0.78

apimachinery

0.59

StoryboardSegue

0.54

 صوتيه

0.54

 utafitiHapana

0.50

stdc

0.45

interopRequire

0.45

 createState

0.44

पया

0.44

ScopeManager

0.42

Activations Density 0.158%

references to numerical values associated with various attributes or characteristics

No Comments

No Known Activations