INDEX

Explanations

elements related to architecture and design

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.17.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

enumii

-0.84

 فريبيس

-0.81

ArrowToggle

-0.79

 kasarigan

-0.79

 createState

-0.78

 pouvoit

-0.77

ISupport

-0.76

 chaude

-0.75

 récents

-0.75

PerformLayout

-0.75

POSITIVE LOGITS

 entire

0.62

 giant

0.62

竟然

0.62

(!)

0.61

 upside

0.58

 edible

0.56

 fake

0.56

竟

0.54

 literally

0.54

居然

0.53

Activations Density 0.390%

elements related to architecture and design

No Comments

No Known Activations