INDEX

Explanations

references to the concept of "flipping" and "molecular" contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 فريبيس

-0.65

NamedQueries

-0.65

 quarto

-0.61

 apartheid

-0.58

extension

-0.57

 Thackeray

-0.57

"?>

-0.57

 Berner

-0.56

 Michelangelo

-0.56

 Tencent

-0.56

POSITIVE LOGITS

 flip

1.09

Mol

1.00

sn

0.95

 Flip

0.95

Flip

0.94

flip

0.93

 molecular

0.93

Mol

0.88

molecular

0.85

MigrationBuilder

0.84

Activations Density 0.067%

references to the concept of "flipping" and "molecular" contexts

No Comments

No Known Activations