INDEX

Explanations

key elements related to changes or developments within a specified context, often focusing on institutional practices or statuses

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

only

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

 становника

-0.49

tvguidetime

-0.43

Życiorys

-0.43

팎

-0.39

Portail

-0.38

Географиясе

-0.38

पया

-0.38

LabelTagHelper

-0.38

colgroup

-0.37

省市镇

-0.36

POSITIVE LOGITS

only

0.57

Only

0.57

 only

0.55

 Only

0.52

 einzigen

0.48

 ONLY

0.48

 BoxFit

0.47

ONLY

0.47

 allora

0.47

 barely

0.46

Activations Density 0.090%

key elements related to changes or developments within a specified context, often focusing on institutional practices or statuses

only

No Comments

No Known Activations