INDEX

Explanations

specific nouns or proper nouns related to individuals and concepts in various contexts

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

names and titles

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

AddTagHelper

-0.39

WebServlet

-0.35

 origines

-0.33

labus

-0.32

 společnost

-0.32

instancetype

-0.31

 cél

-0.31

getContentPane

-0.31

RequestMethod

-0.30

ArgumentParser

-0.30

POSITIVE LOGITS

ValueStyle

0.72

tanleria

0.60

 виправивши

0.56

 nakalista

0.53

disposing

0.52

AndEndTag

0.52

SharedCtor

0.51

Erreferentziak

0.50

 resourceCulture

0.50

BufferException

0.50

Activations Density 0.000%

specific nouns or proper nouns related to individuals and concepts in various contexts

names and titles

No Comments

No Known Activations