INDEX

Explanations

mathematical symbols and notations related to advanced theories

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

Superscripts (often "abc", "ab", etc.)

np_acts-logits-general · gemini-2.0-flash

abc sequences

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

īts

-0.52

 Waterhouse

-0.49

 createState

-0.48

 Quentin

-0.48

 challenges

-0.47

ēr

-0.47

 tunt

-0.47

IsContent

-0.46

weł

-0.46

 Persona

-0.46

POSITIVE LOGITS

ab

1.09

AB

1.06

AB

1.05

abc

1.04

ab

1.00

ABC

0.98

abc

0.95

abcd

0.95

 ABCD

0.94

ABC

0.94

Activations Density 1.382%

mathematical symbols and notations related to advanced theories

Superscripts (often "abc", "ab", etc.)

abc sequences

No Comments

No Known Activations