INDEX

Explanations

phrases related to news articles and journalism

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 challeng

-0.69

 nostalg

-0.65

ItemTracker

-0.64

etheless

-0.63

<|endoftext|>

-0.63

ortium

-0.60

CONCLUS

-0.60

iterranean

-0.60

aturdays

-0.58

edIn

-0.57

POSITIVE LOGITS

Phys

0.81

 Shutterstock

0.74

 Transcript

0.66

 âĢº

0.66

UTERS

0.65

utterstock

0.65

 Hillary

0.65

 Mandatory

0.63

 Caption

0.63

=>

0.62

Activations Density 0.117%

phrases related to news articles and journalism

No Comments

No Known Activations