INDEX

Explanations

occurrences of numbers, particularly those that are two or more digits

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

numbers preceding time or quantity

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

是一种

-0.44

了一种

-0.42

是一種

-0.41

一條

-0.40

MessageTagHelper

-0.39

 وتسجيلات

-0.39

է

-0.38

 isKindOfClass

-0.38

是一个

-0.38

POSITIVE LOGITS

Remaining

0.57

testens

0.56

OrEmpty

0.54

 Remaining

0.54

 остальных

0.51

remaining

0.49

ennan

0.49

omock

0.47

irited

0.47

 ensuing

0.47

Activations Density 0.027%

occurrences of numbers, particularly those that are two or more digits

numbers preceding time or quantity

No Comments

No Known Activations