INDEX

Explanations

phrases and concepts related to accidents, errors, and coincidences

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

accidental error oversight

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

 autorytatywna

-0.87

Autoritní

-0.76

 мәкал

-0.75

SharedDtor

-0.69

 صوتيه

-0.68

Spoljašnje

-0.67

 defaultstate

-0.65

 كومونز

-0.65

Diwedd

-0.65

 beginnetje

-0.64

POSITIVE LOGITS

LayoutInflater

0.44

ly

0.41

 LayoutInflater

0.38

out

0.37

 occurrence

0.33

or

0.33

act

0.32

yet

0.31

dart

0.31

yet

0.31

Activations Density 0.036%

phrases and concepts related to accidents, errors, and coincidences

accidental error oversight

No Comments

No Known Activations