INDEX

Explanations

actions or activities related to personal experiences and self-reflection

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 incorpor

-0.67

 departures

-0.65

 Canaver

-0.64

 Regulatory

-0.62

 Ernst

-0.59

 Advisory

-0.58

Translation

-0.58

çīĪ

-0.57

Reloaded

-0.57

 additions

-0.57

POSITIVE LOGITS

?",

0.73

 shitty

0.71

 drunk

0.70

 crappy

0.69

 ______

0.69

 sweaty

0.68

 toilet

0.67

 stupid

0.67

 boring

0.67

 menstru

0.65

Activations Density 0.442%

actions or activities related to personal experiences and self-reflection

No Comments

No Known Activations