INDEX

Explanations

words related to awards, nominations, and prizes

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

 Anon

-0.81

tes

-0.70

agen

-0.69

 tempo

-0.68

slow

-0.68

ohyd

-0.68

ths

-0.67

Uz

-0.66

gun

-0.65

begin

-0.64

POSITIVE LOGITS

 awarded

1.39

 awards

1.29

 winner

1.27

 nominations

1.24

 winners

1.22

 award

1.20

 laureate

1.18

 Award

1.17

 laure

1.17

Winner

1.15

Activations Density 1.028%

words related to awards, nominations, and prizes

No Comments

No Known Activations