INDEX

Explanations

words related to historical figures and developments, particularly in ancient civilizations

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

 GOODMAN

-0.71

ICAN

-0.67

 IMAGES

-0.64

lihood

-0.61

fet

-0.60

 Halls

-0.60

 Yanuk

-0.60

IONS

-0.57

 Doctrine

-0.57

 Maiden

-0.56

POSITIVE LOGITS

venants

1.11

ingen

1.09

ffe

1.07

asters

1.04

ppers

1.03

founder

1.03

olly

1.00

ven

0.98

ops

0.97

herence

0.97

Activations Density 0.360%

words related to historical figures and developments, particularly in ancient civilizations

No Comments

No Known Activations