INDEX

Explanations

phrases related to instructions or regulations

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

 Aires

-0.56

 abundantly

-0.55

 banner

-0.55

 Corrections

-0.54

 Penal

-0.53

 marginal

-0.52

nu

-0.52

eh

-0.51

SPONSORED

-0.51

 sterling

-0.51

POSITIVE LOGITS

0.95

-$

0.93

usterity

0.93

alog

0.89

lihood

0.84

bsite

0.83

mosp

0.82

etheless

0.80

tenance

0.80

Activations Density 0.612%

phrases related to instructions or regulations

No Comments

No Known Activations