INDEX

Explanations

statistics and data related to different regions and topics

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

 guise

-0.84

 orche

-0.84

 abduct

-0.82

 authentic

-0.82

 hostage

-0.81

 imperson

-0.81

 appropriate

-0.79

 discreet

-0.79

 gaze

-0.79

 inform

-0.78

POSITIVE LOGITS

 Compared

1.33

 Additionally

1.32

 Compare

1.29

 Meanwhile

1.28

 Similarly

1.28

 Overall

1.28

 Conversely

1.27

 Interestingly

1.24

 Including

1.23

 Assuming

1.22

Activations Density 1.547%

statistics and data related to different regions and topics

No Comments

No Known Activations