INDEX

Explanations

phrases related to correctness and criticism in assessments or arguments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 насељу

-0.60

 jLabel

-0.58

 contradic

-0.58

phosa

-0.58

 rospy

-0.56

dération

-0.55

aporation

-0.55

romy

-0.54

 createSprite

-0.54

Gön

-0.54

POSITIVE LOGITS

Personensuche

0.57

 newOwner

0.54

 فريبيس

0.51

 része

0.51

 rightly

0.50

Viitteet

0.48

 HttpNotFound

0.47

zeciw

0.47

 rightfully

0.46

elemField

0.46

Activations Density 0.427%

phrases related to correctness and criticism in assessments or arguments

No Comments

No Known Activations