INDEX

Explanations

phrases indicating the presentation of research findings or results

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

show that

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 يتيمه

-0.50

 FetchType

-0.49

 Normdatei

-0.48

 فريبيس

-0.46

 незавершена

-0.45

 kimse

-0.43

 esternos

-0.43

undefined

-0.42

ieR

-0.42

putnik

-0.42

POSITIVE LOGITS

 demonstrating

0.61

 demonstrate

0.60

 Demonstrate

0.54

 demonstrated

0.53

 demostr

0.52

 demonstrates

0.50

 demostra

0.50

 demuestra

0.50

 demostrar

0.49

 showing

0.48

Activations Density 0.059%

phrases indicating the presentation of research findings or results

show that

No Comments

No Known Activations