INDEX

Explanations

occurrences of the word "agen" and related terms in various contexts, often indicating a specific subject or entity

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_65k/average_l0_73

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ReusableCell

-0.71

ⓧ

-0.68

Erreferentziak

-0.66

 referenties

-0.64

Hochspringen

-0.61

 igång

-0.60

+#+#

-0.60

 Commencez

-0.60

 Seeder

-0.59

 calendriers

-0.58

POSITIVE LOGITS

 sofa

0.71

 sofas

0.68

 laure

0.67

 requ

0.62

əsi

0.61

 tuong

0.61

 couches

0.57

 Sofa

0.57

interop

0.57

 वृ

0.57

Activations Density 0.086%

occurrences of the word "agen" and related terms in various contexts, often indicating a specific subject or entity

No Known Activations