INDEX

Explanations

names of people lamenting or involved in negative activities

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

 EDITION

-0.68

Reloaded

-0.66

 Thumbnails

-0.65

RIS

-0.64

CLASSIFIED

-0.64

AAA

-0.63

 ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½

-0.63

PBS

-0.62

 à¨

-0.62

CLS

-0.60

POSITIVE LOGITS

enson

1.10

jan

0.99

auer

0.94

leck

0.91

aney

0.91

kowski

0.89

laus

0.88

zman

0.87

inson

0.86

linger

0.86

Activations Density 1.789%

names of people lamenting or involved in negative activities

No Comments

No Known Activations