INDEX

Explanations

positive descriptions and experiences related to comfort and pleasantness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

MemoryWarning

-0.66

WithIdentifier

-0.59

nbsp

-0.52

zij

-0.52

asche

-0.51

 oxidizing

-0.51

 berani

-0.51

sph

-0.50

taz

-0.50

/***/

-0.50

POSITIVE LOGITS

 pleasant

1.69

Pleasant

1.53

pleasant

1.48

 Pleasant

1.43

 agradable

1.28

 pleasantly

1.27

 unpleasant

1.22

 agradável

1.16

 agréable

1.14

 piacevole

1.12

Activations Density 0.282%

positive descriptions and experiences related to comfort and pleasantness

No Comments

No Known Activations