INDEX

Explanations

terms related to reports, studies, and their findings in research contexts

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

reporting findings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 гла

-0.32

mp

-0.30

Ј

-0.28

mot

-0.28

 خش

-0.26

 besti

-0.26

↵↵

-0.26

 contain

-0.25

 containing

-0.25

fac

-0.25

POSITIVE LOGITS

HideFlags

0.82

Билгалдахарш

0.79

Datuak

0.73

----</

0.71

 Geſch

0.69

 الرياضيه

0.67

/*---

0.64

 TextInputType

0.63

niſſe

0.62

[@BOS@]

0.62

Activations Density 0.479%

terms related to reports, studies, and their findings in research contexts

reporting findings

No Comments

No Known Activations