INDEX

Explanations

instances of positive and negative test results related to medical conditions

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

 reforms

-0.40

cmds

-0.40

style

-0.38

mjs

-0.38

 StyleSheet

-0.37

Lato

-0.35

LookAnd

-0.35

curr

-0.35

 style

-0.34

AlterField

-0.34

POSITIVE LOGITS

 nakalista

0.70

 pinulongan

0.70

mybatisplus

0.69

PostInfinity

0.59

MLLoader

0.56

posedge

0.56

 Positive

0.54

POSITIVE

0.53

Positive

0.53

istoitu

0.52

Activations Density 0.051%

instances of positive and negative test results related to medical conditions

No Comments

No Known Activations