INDEX

Explanations

information related to news events, security operations, and political issues

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 increasingly

-0.88

 spir

-0.80

 unaff

-0.78

 quir

-0.78

 carving

-0.78

 shifting

-0.77

 ende

-0.76

 casc

-0.76

 shrinking

-0.75

 regener

-0.75

POSITIVE LOGITS

Advertisements

1.55

Discussion

1.42

Sources

1.40

References

1.39

Trivia

1.37

Conclusion

1.35

Official

1.33

Interview

1.32

Posted

1.32

Quotes

1.32

Activations Density 0.514%

information related to news events, security operations, and political issues

No Comments

No Known Activations