INDEX

Explanations

references to legal or political disputes and the individuals involved in them

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 19-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_19/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.19.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

 calendriers

-0.64

 prenn

-0.62

 religieuses

-0.60

UnitTesting

-0.59

 démocr

-0.58

 quelcon

-0.58

 vervol

-0.58

 pauv

-0.57

 épar

-0.57

 förstå

-0.57

POSITIVE LOGITS

 accidentally

0.96

 refused

0.78

 allegedly

0.75

 inadvertently

0.73

 voluntarily

0.73

 attempted

0.73

 accused

0.73

 vandal

0.72

 fainted

0.68

 mistakenly

0.67

Activations Density 0.825%

references to legal or political disputes and the individuals involved in them

No Comments

No Known Activations