INDEX

Explanations

expressions of positivity or approval

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

DllImport

-0.63

ungkinan

-0.60

 downvotes

-0.60

ObjectContext

-0.58

तान

-0.58

zul

-0.56

 tratti

-0.56

Скачать

-0.56

 étoit

-0.56

 sánchez

-0.55

POSITIVE LOGITS

 good

0.86

bad

0.78

 bagus

0.75

 nice

0.72

 טוב

0.69

 Good

0.67

BAD

0.67

good

0.67

 bueno

0.66

Bad

0.65

Activations Density 0.130%

expressions of positivity or approval

No Comments

No Known Activations