INDEX

Explanations

references to journal entries or academic citations

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

The letter "j" (often in math contexts)

np_acts-logits-general · gemini-2.0-flash

variable j in expressions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

بوابة

-0.49

abstractmethod

-0.46

IBar

-0.45

AuthProvider

-0.44

wnątrz

-0.44

Autres

-0.43

 encontraban

-0.39

 rozwój

-0.39

CBA

-0.39

 negó

-0.39

POSITIVE LOGITS

 defaultstate

0.99

0.83

0.81

Jh

0.80

Juf

0.79

JJ

0.77

JJ

0.77

Jaw

0.75

Joc

0.75

Activations Density 0.697%

references to journal entries or academic citations

The letter "j" (often in math contexts)

variable j in expressions

No Comments

No Known Activations