INDEX

Explanations

references to healthcare contexts and connections between various elements in studies or data

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

archaic spellings and prepositions

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 jambes

-0.40

 instead

-0.39

and

-0.38

WriteAttribute

-0.36

 illustrationer

-0.34

fritt

-0.33

 motyw

-0.33

instead

-0.32

 Decke

-0.31

able

-0.31

POSITIVE LOGITS

 Majefty

0.66

脚注の使い方

0.64

UnknownFieldSet

0.60

DockStyle

0.59

ſelf

0.59

 myſelf

0.55

ſelves

0.54

 noDo

0.54

 ſch

0.52

 ―――――

0.52

Activations Density 0.039%

references to healthcare contexts and connections between various elements in studies or data

archaic spellings and prepositions

No Comments

No Known Activations