INDEX

Explanations

tokens that represent various identifiers or codes, potentially related to formatting or sections within a structured document

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

Strings of letters and numbers

np_acts-logits-general · gemini-2.0-flash

venopril, government, ruogosa

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

 виправивши

-0.62

aarrggbb

-0.60

 NSCoder

-0.59

ContentAlignment

-0.58

InstanceState

-0.57

SerializedName

-0.56

findpost

-0.56

astéroïdes

-0.55

Хьажоргаш

-0.55

 שוליים

-0.55

POSITIVE LOGITS

 Majefty

0.64

 reaſon

0.57

 gani

0.54

Autoritní

0.53

 contextLoads

0.51

TOBER

0.50

tober

0.49

ὶ

0.47

 Diſ

0.47

SEGUIR

0.47

Activations Density 0.212%

tokens that represent various identifiers or codes, potentially related to formatting or sections within a structured document

Strings of letters and numbers

venopril, government, ruogosa

No Comments

No Known Activations