INDEX

Explanations

phrases that indicate methodology or manner of action

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 houſe

-0.87

 greateſt

-0.74

 Houſe

-0.74

PyTuple

-0.69

 Diſ

-0.66

Efq

-0.64

 ferien

-0.64

 løpet

-0.64

 ſtate

-0.63

حيان

-0.63

POSITIVE LOGITS

 тому

0.60

 mano

0.59

[@

0.58

EDEFAULT

0.56

 forn

0.56

Donnell

0.56

 kasarigan

0.55

 contro

0.55

upsi

0.55

antMatchers

0.55

Activations Density 0.207%

phrases that indicate methodology or manner of action

No Comments

No Known Activations