INDEX

Explanations

references to genomic and biological data, particularly focusing on structural and quantitative aspects

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

:✨

-0.49

ikian

-0.47

 fufficient

-0.46

 microm

-0.46

 GoogleFonts

-0.46

kao

-0.45

WithIOException

-0.45

ymus

-0.45

penup

-0.45

axiom

-0.45

POSITIVE LOGITS

 resides

0.45

 gdyby

0.45

 financière

0.45

 financieras

0.45

 financeiro

0.43

 arbete

0.43

 reside

0.43

 empleos

0.41

 revolves

0.41

 migrantes

0.40

Activations Density 0.075%

references to genomic and biological data, particularly focusing on structural and quantitative aspects

No Comments

No Known Activations