INDEX

Explanations

words or phrases denoting technical, biological, and methodological descriptions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/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.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 propOrder

-1.03

HtmlAttribute

-0.96

出版年

-0.95

 nahilalakip

-0.93

новниш

-0.91

LEncoder

-0.87

 jScrollPane

-0.86

MigrationBuilder

-0.86

 виправивши

-0.85

脚注の使い方

-0.83

POSITIVE LOGITS

 gezien

0.88

 speaking

0.86

 parlant

0.86

 hablando

0.85

 parlando

0.76

 gesehen

0.74

 falando

0.73

speaking

0.66

ically

0.66

 Speaking

0.61

Activations Density 0.221%

words or phrases denoting technical, biological, and methodological descriptions

No Comments

No Known Activations