INDEX

Explanations

instances of text that start with a special token or denote the beginning of a new segment, including the initial token of data streams

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

kegaard

-0.52

 degeneracy

-0.52

 demurrer

-0.49

ArgsConstructor

-0.45

uggest

-0.45

Petite

-0.45

canestro

-0.44

 tetrach

-0.44

裁判

-0.44

 suggestion

-0.44

POSITIVE LOGITS

 '\\;'

0.92

 juiste

0.81

存于互联网档案馆

0.79

 فريبيس

0.79

@"/

0.78

Архівовано

0.77

AntiForgeryToken

0.74

Personendaten

0.73

 appropriate

0.72

 correct

0.72

Activations Density 0.199%

instances of text that start with a special token or denote the beginning of a new segment, including the initial token of data streams

No Comments

No Known Activations