INDEX

Explanations

mathematical terms or notations related to functions and operators

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

Mathematical notation: "mathrm"

np_acts-logits-general · gemini-2.0-flash

`mathrm` followed by brace

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 AssemblyVersion

-1.02

OGND

-1.00

windowFixed

-0.99

StoryboardSegue

-0.99

setVerticalGroup

-0.97

Panamoan

-0.91

AsUp

-0.90

 typelib

-0.89

 ExecuteAsync

-0.89

 propOrder

-0.88

POSITIVE LOGITS

{~

0.97

mathrm

0.85

0.59

Rams

0.55

 Rams

0.53

 $('<

0.52

 Hewitt

0.49

 Korn

0.49

 Milli

0.48

Activations Density 0.023%

mathematical terms or notations related to functions and operators

Mathematical notation: "mathrm"

`mathrm` followed by brace

No Comments

No Known Activations