INDEX

Explanations

adverbs indicating probability or expectation

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

iya

-0.75

bern

-0.75

issy

-0.72

ortmund

-0.70

enberg

-0.70

ilian

-0.69

aeus

-0.68

osi

-0.68

elight

-0.67

aan

-0.66

POSITIVE LOGITS

 underest

0.80

 underestimate

0.78

 overest

0.75

 underestimated

0.74

won

0.74

 misunder

0.73

won

0.72

 influenced

0.70

 wont

0.69

 exagger

0.69

Activations Density 0.423%

adverbs indicating probability or expectation

No Comments

No Known Activations