INDEX

Explanations

locations, especially related to medical facilities

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

 Marketable

-0.73

thin

-0.72

CHAT

-0.70

ãģĦ

-0.66

lass

-0.65

lust

-0.64

alam

-0.64

grain

-0.62

estamp

-0.62

lasses

-0.62

POSITIVE LOGITS

 entrances

1.01

 premises

0.99

 entrance

0.98

 located

0.96

 occupancy

0.93

 doors

0.93

 operated

0.92

 attendant

0.89

 staffed

0.86

 renovations

0.86

Activations Density 1.513%

locations, especially related to medical facilities

No Comments

No Known Activations