INDEX

Explanations

conditional statements or recommendations

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

!*\

-0.62

ftagPool

-0.59

 EconPapers

-0.55

ImageContext

-0.53

IPAC

-0.52

valentino

-0.52

ToMany

-0.52

 demais

-0.51

DECREF

-0.50

 abstrato

-0.50

POSITIVE LOGITS

should

1.00

 Should

0.88

Should

0.87

 should

0.85

ShouldBe

0.80

 SHOULD

0.79

hould

0.74

 debería

0.58

 Sollte

0.57

 deberían

0.56

Activations Density 0.001%

conditional statements or recommendations

No Comments

No Known Activations