INDEX

Explanations

overall assessments or summaries

oai_token-act-pair · gpt-3.5-turbo

statements or summaries of overall assessments

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.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

 bound

-0.74

 forbidden

-0.73

 demanding

-0.68

Pis

-0.68

 subpoena

-0.67

 beck

-0.64

 forbid

-0.63

 pleading

-0.63

dep

-0.63

cod

-0.62

POSITIVE LOGITS

Overall

3.43

 Overall

3.18

 overall

1.89

Conclusion

1.46

Ultimately

1.30

Total

1.29

Compared

1.26

Bottom

1.25

 Lastly

1.22

Recommend

1.20

Activations Density 0.024%

overall assessments or summaries

statements or summaries of overall assessments

No Comments

No Known Activations