INDEX

Explanations

terms related to statistical analysis and decision-making based on data and probabilities

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

 aven

-0.79

 gunfire

-0.72

 grip

-0.71

eteen

-0.71

 defiant

-0.70

 quir

-0.70

 mourn

-0.69

 earthqu

-0.69

 unbeaten

-0.69

 amused

-0.68

POSITIVE LOGITS

 Therefore

1.92

 Consequently

1.79

 Hence

1.75

 Thus

1.73

 Furthermore

1.69

 Additionally

1.68

 Ideally

1.67

 Conversely

1.66

 Moreover

1.63

 Whereas

1.61

Activations Density 3.023%

terms related to statistical analysis and decision-making based on data and probabilities

No Comments

No Known Activations