INDEX

Explanations

mentions of specific locations, such as towns and cities

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

isSpecialOrderable

-0.77

 psychosis

-0.72

izoph

-0.72

gression

-0.71

 unanimous

-0.70

 streaks

-0.68

 flashbacks

-0.67

 Trayvon

-0.67

 replay

-0.67

 definitively

-0.66

POSITIVE LOGITS

 manufacturing

1.36

 factories

1.34

 manufactures

1.28

 industries

1.25

 manufacture

1.24

 shops

1.24

 artisan

1.22

 mills

1.21

 suppliers

1.21

 commodities

1.17

Activations Density 2.783%

mentions of specific locations, such as towns and cities

No Comments

No Known Activations