INDEX

Explanations

phrases related to legal actions and court rulings

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

ription

-0.74

ontent

-0.73

-0.72

nels

-0.71

 weap

-0.69

 Horus

-0.69

ufact

-0.69

cov

-0.68

 volunteers

-0.66

 broch

-0.66

POSITIVE LOGITS

 justices

1.33

 upholding

1.19

 rulings

1.17

 Scalia

1.15

 ruling

1.11

 overturn

1.10

 upheld

1.05

Justice

1.04

Judge

1.03

 sided

1.03

Activations Density 1.158%

phrases related to legal actions and court rulings

No Comments

No Known Activations