INDEX

Explanations

textual features associated with instructions or procedural information

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

abled

-0.67

rams

-0.66

paralle

-0.66

nz

-0.63

sports

-0.62

sung

-0.62

ral

-0.62

bm

-0.61

far

-0.60

Share

-0.60

POSITIVE LOGITS

 introdu

1.02

 basics

0.85

 Introduction

0.79

 obligatory

0.75

 impressions

0.74

rontal

0.73

 initialize

0.72

 obvious

0.71

 hello

0.71

 congratulations

0.69

Activations Density 1.034%

textual features associated with instructions or procedural information

No Comments

No Known Activations