INDEX

Explanations

phrases that indicate causality or reasoning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

NOPQRST

-0.50

شاف

-0.48

)))),

-0.47

southeast

-0.47

 зачем

-0.47

vědět

-0.47

 fous

-0.46

katapos

-0.46

unächst

-0.46

 gesprek

-0.46

POSITIVE LOGITS

Due

0.98

 Debido

0.88

Because

0.86

Due

0.86

Debido

0.84

 Aufgrund

0.82

due

0.82

ůli

0.81

 Because

0.79

 Owing

0.79

Activations Density 0.036%

phrases that indicate causality or reasoning

No Comments

No Known Activations