INDEX

Explanations

phrases expressing opinions and judgments about performance or events

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

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

 ivelany

-0.56

 saites

-0.51

TokenNameLPAREN

-0.49

tagHelperRunner

-0.48

uxxxx

-0.47

DockStyle

-0.47

 utafitiHapana

-0.46

brigens

-0.45

featureID

-0.45

الحياه

-0.45

POSITIVE LOGITS

 replay

0.44

did

0.41

 relive

0.38

 tört

0.38

 replays

0.37

did

0.37

detectChanges

0.37

振り返

0.37

 slept

0.37

 endforeach

0.37

Activations Density 0.027%

phrases expressing opinions and judgments about performance or events

No Comments

No Known Activations