INDEX

Explanations

names of locations, such as cities and countries

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

 Lauder

-0.72

xual

-0.65

ertodd

-0.62

epad

-0.61

 Ital

-0.60

yang

-0.60

 Slate

-0.59

roups

-0.58

cair

-0.57

nels

-0.56

POSITIVE LOGITS

 largest

1.05

 tallest

1.02

 longest

0.96

 premiere

0.95

 foremost

0.91

 greatest

0.90

 Greatest

0.90

 biggest

0.89

 smallest

0.89

 oldest

0.89

Activations Density 0.087%

names of locations, such as cities and countries

No Comments

No Known Activations