INDEX

Explanations

the word "a" used as an article

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

twimg

-0.57

ább

-0.55

OSSARY

-0.54

quiler

-0.52

 nonUne

-0.52

 voilà

-0.50

astéroïdes

-0.50

Demografía

-0.49

genau

-0.49

อะไร

-0.49

POSITIVE LOGITS

 jako

0.85

作為

0.80

jako

0.80

Jako

0.79

作为一个

0.75

作为

0.73

 autorytatywna

0.73

 як

0.71

 sebagai

0.67

 bilang

0.67

Activations Density 0.121%

the word "a" used as an article

No Comments

No Known Activations