INDEX

Explanations

references to "hill" in various contexts

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

"Hill" or "hills"

np_acts-logits-general · gemini-2.0-flash

Hill

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Lightboxes

-0.92

 CiNii

-0.74

ngdoc

-0.69

 AssemblyCulture

-0.67

bibfield

-0.66

GEBURTSDATUM

-0.65

 hObject

-0.65

LookAnd

-0.64

WireFormatLite

-0.64

ellas

-0.63

POSITIVE LOGITS

top

0.81

billy

0.74

 hill

0.69

 hills

0.68

crest

0.65

 Hills

0.62

side

0.61

 Hill

0.60

ock

0.60

Hil

0.58

Activations Density 0.066%

references to "hill" in various contexts

"Hill" or "hills"

Hill

No Comments

No Known Activations