INDEX

Explanations

phrases related to articles or pieces of news

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

 suppose

-0.68

 conced

-0.67

pport

-0.66

 owing

-0.65

 such

-0.64

indal

-0.62

 secondly

-0.61

 inflicting

-0.60

lia

-0.60

 inhib

-0.60

POSITIVE LOGITS

atre

1.02

 latest

0.98

odore

0.93

ater

0.88

 widest

0.85

resa

0.84

 entirety

0.83

 following

0.77

 entire

0.76

 Conversation

0.76

Activations Density 0.278%

phrases related to articles or pieces of news

No Comments

No Known Activations