INDEX

Explanations

terms related to support and contributions in scientific or analytical contexts

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

"[something] lends itself" or "[something] gives [something]"

np_acts-logits-general · gemini-2.0-flash

enables or lends itself

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

stdc

-0.76

makeConstraints

-0.70

AntiForgeryToken

-0.70

enumii

-0.68

expandindo

-0.67

myself

-0.66

 Myself

-0.66

 يتيمه

-0.64

 surla

-0.63

AfterClass

-0.63

POSITIVE LOGITS

us

0.76

 itself

0.71

 quite

0.67

 much

0.59

 nicely

0.58

 собой

0.58

 pretty

0.58

 nothing

0.55

什么意思

0.54

Activations Density 0.831%

terms related to support and contributions in scientific or analytical contexts

"[something] lends itself" or "[something] gives [something]"

enables or lends itself

No Comments

No Known Activations