INDEX

Explanations

numeric values and their frequencies in a medical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_mlp_out

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

 purpoſe

-0.84

 myſelf

-0.83

 ſtate

-0.82

 Monfieur

-0.78

 preſent

-0.77

 Jefus

-0.75

 ſta

-0.75

 ſtand

-0.74

 ſhe

-0.74

 houſe

-0.73

POSITIVE LOGITS

 autorytatywna

0.59

 BoxFit

0.52

 years

0.50

شهاد

0.47

to

0.47

pa

0.46

 business

0.46

 internetowa

0.46

 feeds

0.45

0.44

Activations Density 0.378%

numeric values and their frequencies in a medical context

No Comments

No Known Activations