INDEX

Explanations

expressions of help and appreciation

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

Requests followed by "would be appreciated"

np_acts-logits-general · gemini-2.0-flash

would be appreciated

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 enfans

-0.64

SharedDtor

-0.64

Tikang

-0.62

PageContext

-0.59

 leçons

-0.58

 sauvages

-0.57

 coraggio

-0.56

 devriez

-0.56

 noires

-0.56

 BoxDecoration

-0.54

POSITIVE LOGITS

 welcome

0.95

 welcomed

0.89

 preferred

0.89

 appreciated

0.85

 nice

0.73

 gratefully

0.71

preferred

0.70

 WELCOME

0.69

 greatly

0.69

MemoryWarning

0.68

Activations Density 0.158%

expressions of help and appreciation

Requests followed by "would be appreciated"

would be appreciated

No Comments

No Known Activations