INDEX

Explanations

phrases related to multi-part projects

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

ubs

-0.82

wolves

-0.75

ampires

-0.74

osponsors

-0.70

mia

-0.68

bara

-0.68

uca

-0.68

DERR

-0.67

artifacts

-0.67

ateurs

-0.67

POSITIVE LOGITS

 series

0.81

 combo

0.79

 puzzle

0.77

 investigative

0.74

 saga

0.73

 sequential

0.73

 cumulative

0.73

 strategy

0.72

 affair

0.72

 formula

0.72

Activations Density 0.045%

phrases related to multi-part projects

No Comments

No Known Activations