INDEX

Explanations

technical codes, including hashes and IDs

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

 checklist

-0.68

DonaldTrump

-0.66

 bisc

-0.66

 newsletters

-0.63

itsch

-0.63

 totality

-0.62

 signs

-0.62

 mediation

-0.61

 experien

-0.60

 contrace

-0.58

POSITIVE LOGITS

"}

0.94

AAAAAAAA

0.94

}\

0.82

wr

0.81

[_

0.80

\",

0.80

\"

0.79

usercontent

0.79

"},

0.78

:\

0.77

Activations Density 0.703%

technical codes, including hashes and IDs

No Comments

No Known Activations