INDEX

Explanations

technical terms and information related to electronics and circuits

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

 ÂŃ

-1.07

ÂŃ

-0.84

 Rodham

-0.73

kefeller

-0.71

 amid

-0.69

iberal

-0.68

 imper

-0.67

AFP

-0.67

,''

-0.66

 constitutional

-0.66

POSITIVE LOGITS

;)

1.13

 devs

1.11

 alot

1.11

 kinda

1.10

:)

1.10

 haha

1.06

 ðŁĻĤ

1.02

 youtube

1.00

 english

0.97

 Youtube

0.96

Activations Density 6.982%

technical terms and information related to electronics and circuits

No Comments

No Known Activations