INDEX

Explanations

references to railway lines or railroads

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

arios

-0.79

iaries

-0.79

oln

-0.75

ividual

-0.73

vironment

-0.73

osi

-0.73

iances

-0.71

isters

-0.70

arial

-0.70

eanor

-0.70

POSITIVE LOGITS

roads

1.17

 routes

0.89

 stations

0.88

road

0.86

 railway

0.86

 station

0.85

 railways

0.83

 locom

0.82

ways

0.81

 corridors

0.81

Activations Density 0.013%

references to railway lines or railroads

No Comments

No Known Activations