INDEX

Explanations

terms related to medical conditions and their treatments

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

 Autorisations

-0.49

énario

-0.47

 cejas

-0.47

 fratelli

-0.46

สอบ

-0.43

 nakalista

-0.43

 Anfrage

-0.42

plaatst

-0.42

 braccia

-0.42

kijken

-0.41

POSITIVE LOGITS

хьтан

0.56

 lenker

0.51

kwds

0.48

 Савезне

0.47

cav

0.46

aka

0.44

Vidite

0.42

Hallo

0.42

 coro

0.41

俗

0.41

Activations Density 0.131%

terms related to medical conditions and their treatments

No Comments

No Known Activations