INDEX

Explanations

references to routine activities and responsibilities

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

entity

-0.87

 precedent

-0.80

canon

-0.78

 definitively

-0.78

DNA

-0.76

worthiness

-0.76

 unequivocally

-0.76

 predictive

-0.74

 flaws

-0.73

 ISBN

-0.73

POSITIVE LOGITS

 meals

1.48

 snacks

1.41

 dinners

1.33

 supper

1.32

 chores

1.29

 breakfast

1.28

 dinner

1.25

 snack

1.25

 meal

1.24

 lunch

1.23

Activations Density 6.752%

references to routine activities and responsibilities

No Comments

No Known Activations