INDEX

Explanations

mathematical or scientific terms and concepts

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

llan

-0.71

riad

-0.68

committee

-0.62

 sensing

-0.61

scenes

-0.60

iculty

-0.60

 Beaut

-0.60

olean

-0.60

olicy

-0.60

rely

-0.58

POSITIVE LOGITS

unciation

1.51

unci

1.36

ounced

1.29

ihilation

1.22

ouncing

1.20

ounces

1.18

ational

1.17

atural

1.13

ials

1.10

ounce

1.08

Activations Density 1.097%

mathematical or scientific terms and concepts

No Comments

No Known Activations