INDEX

Explanations

specific references to countries and political figures

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

 besides

-0.84

thood

-0.74

adays

-0.71

atoon

-0.70

 interacted

-0.70

 namely

-0.70

poon

-0.69

 owing

-0.68

LOD

-0.67

 beforehand

-0.66

POSITIVE LOGITS

 slightest

1.21

 nation

1.20

ocracy

1.13

ocratic

1.13

resa

1.09

 electorate

1.07

 Clintons

1.06

 greatest

1.04

 presidency

1.02

 United

1.02

Activations Density 0.765%

specific references to countries and political figures

No Comments

No Known Activations