INDEX

Explanations

references to authors and their contributions in various fields including law, science, and history

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

Preceding nouns, often capitalized

np_acts-logits-general · gemini-2.0-flash

versus without

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 purpoſe

-0.94

 للاسماء

-0.78

 pleaſure

-0.76

Personendaten

-0.71

 greateſt

-0.71

 theils

-0.69

 reaſon

-0.68

 незавершена

-0.68

 ſmall

-0.68

 houſe

-0.67

POSITIVE LOGITS

 without

0.93

 Without

0.89

 versus

0.85

vs

0.83

without

0.81

 WITHOUT

0.80

Without

0.79

in

0.71

 beyond

0.71

 ilman

0.70

Activations Density 0.347%

references to authors and their contributions in various fields including law, science, and history

Preceding nouns, often capitalized

versus without

No Comments

No Known Activations