INDEX

Explanations

relationships between different concepts or categories

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

Followed by "one"

np_acts-logits-general · gemini-2.0-flash

than one

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 betweenstory

-0.76

tinyos

-0.62

 للمعارف

-0.60

 Мексичка

-0.57

الحياه

-0.54

thoven

-0.53

enterOuterAlt

-0.52

OrNil

-0.52



-0.50

 CreateTagHelper

-0.49

POSITIVE LOGITS

one

4.00

one

2.54

 één

2.25

uno

2.11

 unul

2.06

 одного

2.06

 یکی

2.02

 один

2.02

 אחד

2.01

One

1.99

Activations Density 2.387%

relationships between different concepts or categories

Followed by "one"

than one

No Comments

No Known Activations