INDEX

Explanations

instances where reasons or justifications are being stated

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

ształ

-0.47

 utafitiHapana

-0.47

Vil

-0.44

arim

-0.42

 оп

-0.42

Am

-0.42

过

-0.42

過

-0.41

An

-0.39

']")

-0.39

POSITIVE LOGITS

verwijspagina

0.94

RegistryLite

0.93

 perché

0.79

 kasarigan

0.78

 porque

0.77

porque

0.76

 because

0.75

之所以

0.74

 becauſe

0.74

 Çünkü

0.72

Activations Density 0.371%

instances where reasons or justifications are being stated

No Comments

No Known Activations