INDEX

Explanations

numerical values expressed in various formats

oai_token-act-pair · gpt-3.5-turbo

numerical data, particularly in financial and statistical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 likeness

-0.67

 ambassadors

-0.67

 domest

-0.66

 pseudonym

-0.64

 conserv

-0.62

cffffcc

-0.61

ement

-0.61

 diplomatic

-0.60

 habit

-0.59

think

-0.59

POSITIVE LOGITS

 TOTAL

1.03

 Lastly

0.95

 Finally

0.82

ibaba

0.80

 Ibid

0.80

 Total

0.80

Total

0.76

%);

0.76

 Difference

0.75

Conclusion

0.74

Activations Density 0.223%

numerical values expressed in various formats

numerical data, particularly in financial and statistical contexts

No Comments

No Known Activations