INDEX

Explanations

terms associated with legal terminology and programming

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 alimentaire

-0.73

 alimentaires

-0.72

iettivo

-0.67

atguigu

-0.66

 dramatique

-0.63

ⓧ

-0.62

rativa

-0.60

Efq

-0.60

 traités

-0.59

alright

-0.59

POSITIVE LOGITS

 оригіналу

0.67

0.58

UnusedPrivate

0.54

rest

0.50

 Benjamin

0.48

Benjamin

0.45

撇

0.44

 beans

0.44

 carrier

0.44

 singular

0.43

Activations Density 2.180%

terms associated with legal terminology and programming

No Comments

No Known Activations