INDEX

Explanations

names of people and places from a specific context or storyline

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

iP

-0.99

 antip

-0.90

 indo

-0.87

IMP

-0.86

 hemor

-0.86

 rapt

-0.85

 transf

-0.85

INF

-0.84

Þ

-0.84

ASP

-0.82

POSITIVE LOGITS

ley

1.59

leys

1.26

LEY

1.21

lett

1.18

ington

1.12

lee

1.11

sey

1.10

rette

1.10

ton

1.09

nell

1.08

Activations Density 0.107%

names of people and places from a specific context or storyline

No Comments

No Known Activations