INDEX

Explanations

references to specific features or options within a technical context

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

accessories and optional extras

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

 Lokales

-0.45

ⓧ

-0.38

 Waray

-0.38

 dealing

-0.37

 compelling

-0.36

发表于

-0.35

ArgsConstructor

-0.35

 quite

-0.35

nullptr

-0.33

 allAfrica

-0.32

POSITIVE LOGITS

 ब्रेकडाउन

0.64

 المعيارى

0.56

 fubject

0.53

Билгалдахарш

0.52

pherals

0.51

Jeografia

0.49

complexContent

0.49

 leſs

0.48

таратура

0.48

schwar

0.47

Activations Density 0.012%

references to specific features or options within a technical context

accessories and optional extras

No Comments

No Known Activations