INDEX

Explanations

phrases related to specific locations or environments

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

apult

-0.77

advertisement

-0.67

ÂłÂłÂłÂłÂłÂłÂłÂł

-0.66

termin

-0.63

mask

-0.63

Pwr

-0.62

ME

-0.60

unes

-0.59

fml

-0.58

POSITIVE LOGITS

upon

1.51

soever

1.05

abouts

0.99

fore

0.89

ver

0.78

 they

0.74

 users

0.72

holders

0.69

we

0.69

 temperatures

0.68

Activations Density 2.555%

phrases related to specific locations or environments

No Comments

No Known Activations