INDEX

Explanations

references to the "Linotype" system and its related terminology

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

ValueStyle

-0.45

 Compro

-0.43

awe

-0.41

paw

-0.40

gtest

-0.40

 пого

-0.40

 Hobson

-0.39

pata

-0.39

ктей

-0.39

[]={

-0.38

POSITIVE LOGITS

Lin

0.93

Lin

0.93

lin

0.82

LIN

0.81

 Linde

0.79

 Lind

0.79

Lind

0.79

lin

0.78

 Liner

0.75

 Linz

0.75

Activations Density 0.015%

references to the "Linotype" system and its related terminology

No Comments

No Known Activations