INDEX

Explanations

text related to education, academic studies, research, and technical terms

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 describ

-0.85

 neighb

-0.81

 carving

-0.80

 tackling

-0.77

 destro

-0.77

 diving

-0.76

 incent

-0.76

 raft

-0.76

 sculpt

-0.75

 prud

-0.74

POSITIVE LOGITS

Advertisements

1.30

Copyright

1.23

Â©

1.17

âĿ

1.15

WHERE

1.15

Posted

1.14

Article

1.14

If

1.13

Therefore

1.13

Join

1.13

Activations Density 0.088%

text related to education, academic studies, research, and technical terms

No Comments

No Known Activations