INDEX

Explanations

declarations of integer and long data types in code

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

numeric types and object ids

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

PLAC

-0.42

 over

-0.39

 sợi

-0.39

ãi

-0.39

底

-0.38

 lucky

-0.38

ert

-0.37

stays

-0.36

 almost

-0.36

getitem

-0.36

POSITIVE LOGITS

 Normdatei

0.61

 AssemblyTitle

0.61

 '\\;'

0.60

AutoField

0.54

parsedMessage

0.53

bibfield

0.51

OGND

0.50

bibinfo

0.50

TagMode

0.47

flashdata

0.47

Activations Density 0.001%

declarations of integer and long data types in code

numeric types and object ids

No Comments

No Known Activations