INDEX

Explanations

references to specific categories of items or subjects, particularly in the context of healthcare and entertainment

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

medical, biological, and chemical terms

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

featureID

-0.67

XDECREF

-0.44

UserScript

-0.42

GTCX

-0.41

tagext

-0.41

 noDo

-0.39

writerow

-0.39

 المعيارى

-0.38

 canst

-0.38

ungsbedingungen

-0.37

POSITIVE LOGITS

fromnode

0.47

rillas

0.45

 Polity

0.44

 <=",

0.42

ally

0.40

uxxxx

0.40

culares

0.40

pondre

0.39

 Schwarz

0.39

lovakia

0.39

Activations Density 0.097%

references to specific categories of items or subjects, particularly in the context of healthcare and entertainment

medical, biological, and chemical terms

No Comments

No Known Activations