INDEX

Explanations

the word "evidence" and its related phrases indicating the presence or quality of evidence in a context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 14-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_14/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.14.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

rrggbb

-0.65

mobileqq

-0.61

 transférez

-0.59

 AssemblyCompany

-0.59

 princely

-0.57

 pasaran

-0.54

IKit

-0.54

 useRouter

-0.53

rural

-0.53

 igång

-0.53

POSITIVE LOGITS

 evidence

1.46

 Evidence

1.35

Evidence

1.30

evidence

1.27

 EVIDENCE

1.20

 evidencia

1.01

 Evid

0.96

 evidences

0.93

 evid

0.88

 eviden

0.86

Activations Density 0.153%

the word "evidence" and its related phrases indicating the presence or quality of evidence in a context

No Comments

No Known Activations