INDEX

Explanations

technical terms related to programming, such as stack operations and software loop parameters

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

 Canaver

-0.89

 Journalism

-0.76

 politic

-0.72

 Architects

-0.69

 renaissance

-0.66

©¶æ¥µ

-0.65

 pastors

-0.64

 Prohibition

-0.62

çīĪ

-0.62

 Patreon

-0.61

POSITIVE LOGITS

 predetermined

0.84

 desired

0.77

 stimuli

0.76

 cursor

0.73

 incoming

0.72

or

0.70

 milliseconds

0.68

 objects

0.68

 inputs

0.68

 increments

0.67

Activations Density 5.328%

technical terms related to programming, such as stack operations and software loop parameters

No Comments

No Known Activations