INDEX

Explanations

assertions and claims about the nature of arguments and their validity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 jScrollPane

-0.52

ніципалі

-0.51

rrggbb

-0.51

ustra

-0.49

 organized

-0.46

ALO

-0.45

mich

-0.45

ACM

-0.44

forall

-0.44

Modi

-0.43

POSITIVE LOGITS

MessageState

0.80

følgelig

0.76

EndContext

0.76

 Normdatei

0.74

enumi

0.73

 religieuses

0.72

 تانيه

0.71

 Walkover

0.71

argout

0.71

verständlich

0.70

Activations Density 0.326%

assertions and claims about the nature of arguments and their validity

No Comments

No Known Activations