INDEX

Explanations

references to personal pronouns and the verb "to be" in various forms

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

ſelf

-1.00

 Majefty

-0.94

ſelves

-0.93

 متعلقه

-0.88

expandindo

-0.88

Efq

-0.86

 Monfieur

-0.85

 myſelf

-0.85

IContainer

-0.85

 ſtate

-0.85

POSITIVE LOGITS

And

0.70

 zwar

0.69

 certainly

0.64

 also

0.62

Dabei

0.58

 especially

0.57

かつ

0.57

And

0.54

 daarbij

0.53

AND

0.53

Activations Density 0.032%

references to personal pronouns and the verb "to be" in various forms

No Comments

No Known Activations