INDEX

Explanations

words related to professional terms or entities

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

ãĤ¼ãĤ¦ãĤ¹

-0.84

ĸļ

-0.73

 straw

-0.72

 needles

-0.71

 Sins

-0.66

 Twain

-0.66

 ponds

-0.65

 spoon

-0.65

 spears

-0.64

owship

-0.64

POSITIVE LOGITS

digy

1.57

blems

1.30

gression

1.28

ceed

1.26

verbs

1.24

spective

1.21

secution

1.19

clamation

1.19

posal

1.16

actively

1.15

Activations Density 0.382%

words related to professional terms or entities

No Comments

No Known Activations