INDEX

Explanations

proper nouns or names

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

ILLE

-0.82

urses

-0.68

abama

-0.65

ITIES

-0.65

scill

-0.63

imation

-0.61

izabeth

-0.61

udder

-0.61

ught

-0.61

ITY

-0.60

POSITIVE LOGITS

 Twain

1.25

eting

1.10

eters

1.09

 Zuckerberg

1.08

ipl

0.99

down

0.98

manship

0.96

erness

0.92

owitz

0.91

edly

0.91

Activations Density 0.642%

proper nouns or names

No Comments

No Known Activations