INDEX

Explanations

concepts related to abstraction and generalization

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

AsUp

-0.66

ContentAlignment

-0.64

 smtplib

-0.63

basicConfig

-0.58

 kasarigan

-0.57

getColumnIndex

-0.57

sidemargin

-0.57

 psycopg

-0.56

ंदीखरीदारी

-0.56

Erreferentziak

-0.55

POSITIVE LOGITS

 فريبيس

0.60

 passive

0.52

StructEnd

0.52

 <>",

0.48

 merely

0.48

 Passive

0.47

Passive

0.47

stag

0.46

 Einzelnen

0.46

 guard

0.46

Activations Density 0.446%

concepts related to abstraction and generalization

No Comments

No Known Activations