INDEX

Explanations

references to content or data that is clearly identified as being above a certain threshold or standard

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

Portail

-0.83

 kasarigan

-0.77

riwal

-0.69

 Gilles

-0.65

😭😭

-0.61

jenigen

-0.61

Rolf

-0.61

ClickListener

-0.61

cal

-0.60

man

-0.60

POSITIVE LOGITS

 ABOVE

1.58

ABOVE

1.56

 Above

1.44

Above

1.39

above

1.32

 BELOW

1.30

 above

1.27

bove

1.26

 Below

1.17

dessus

1.15

Activations Density 0.095%

references to content or data that is clearly identified as being above a certain threshold or standard

No Comments

No Known Activations