INDEX

Explanations

occurrences of the term "word"

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

The token "word"

np_acts-logits-general · gemini-2.0-flash

word of mouth or processor

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

‍♂️

-0.70

atico

-0.63

]-->

-0.61

subplots

-0.61

veratrol

-0.56

@[

-0.55

yní

-0.55

‍♀️

-0.54

gdx

-0.52

 विश्वसनीयता

-0.52

POSITIVE LOGITS

play

0.79

mith

0.69

 processor

0.66

processing

0.64

robe

0.63

smith

0.62

count

0.62

 klucz

0.61

 count

0.60

 Italijani

0.59

Activations Density 0.093%

occurrences of the term "word"

The token "word"

word of mouth or processor

No Comments

No Known Activations