INDEX

Explanations

sequences of underscores or similar repetitive patterns, often used in programming or markup languages

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

Tokens followed by "u" or "be"

np_acts-logits-general · gemini-2.0-flash

u and be types

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.24.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

hoeddwyd

-0.66

IVEREF

-0.65

bootstrapcdn

-0.61

fromnode

-0.59

parsedMessage

-0.58

StoreMessageInfo

-0.57

makeText

-0.54

 OFDb

-0.54

addCriterion

-0.54

adget

-0.50

POSITIVE LOGITS

tagHelperRunner

0.93

Pyx

0.72

pyx

0.68

 Inscrivez

0.62

tagHelper

0.54

attribute

0.52

mû

0.51

 generazione

0.51

webpack

0.50

WEBPACK

0.49

Activations Density 0.126%

sequences of underscores or similar repetitive patterns, often used in programming or markup languages

Tokens followed by "u" or "be"

__u and __be types

No Comments

No Known Activations

u and be types