INDEX

Explanations

phrases that discuss biological mechanisms and relationships, particularly in medical or scientific contexts

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

potential actions or states

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

IUrlHelper

-0.59

RenderAtEndOf

-0.59

 AssemblyCulture

-0.57

ंदीखरीदारी

-0.57

MessageOf

-0.56

المناصب

-0.55

 '\\;'

-0.52

 للمعارف

-0.52

 OMITBAD

-0.52

 surla

-0.49

POSITIVE LOGITS

 generally

0.48

withIdentifier

0.42

is

0.40

Assured

0.39

 attracts

0.38

 only

0.37

 seems

0.37

 basically

0.37

 undoubtedly

0.36

ondheim

0.36

Activations Density 0.138%

phrases that discuss biological mechanisms and relationships, particularly in medical or scientific contexts

potential actions or states

No Comments

No Known Activations