INDEX

Explanations

phrases related to technology and medical terms

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

 Onion

-0.69

 Wick

-0.68

ĸļ

-0.68

çĦ

-0.67

 offence

-0.66

 Sparks

-0.66

 Rabbit

-0.65

 Luck

-0.64

WARD

-0.61

Eck

-0.60

POSITIVE LOGITS

opter

1.05

axis

1.02

ilic

1.00

ically

0.99

etric

0.98

olic

0.97

omic

0.97

itic

0.97

actic

0.96

oidal

0.96

Activations Density 0.970%

phrases related to technology and medical terms

No Comments

No Known Activations