INDEX

Explanations

emphasized expressions of opinion and evaluation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 8-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_8/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.8.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

 linkovi

-0.74

InitVars

-0.71

 henvisninger

-0.70

 initComponents

-0.69

sidemargin

-0.69

 automaticamente

-0.65

SharedCtor

-0.64

Бахар

-0.64

 متعلقه

-0.63

 automáticamente

-0.63

POSITIVE LOGITS

 really

1.02

 truly

1.00

 REALLY

0.95

 realy

0.94

really

0.93

truly

0.90

Truly

0.87

 Really

0.84

 Truly

0.84

 virkelig

0.82

Activations Density 0.121%

emphasized expressions of opinion and evaluation

No Comments

No Known Activations