INDEX

Explanations

names of places or people that include specific letter sequences

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

yip

-0.61

 independents

-0.58

 srfAttach

-0.57

flight

-0.56

 karma

-0.56

acebook

-0.55

 Liberals

-0.54

 Grad

-0.54

pedia

-0.54

Buk

-0.54

POSITIVE LOGITS

roe

0.73

ember

0.71

ethy

0.67

agement

0.66

ĸļ

0.65

vre

0.65

itchie

0.65

aney

0.64

oyer

0.64

ciating

0.62

Activations Density 1.033%

names of places or people that include specific letter sequences

No Comments

No Known Activations