INDEX

Explanations

sentences that end with a positive affirmation or appreciation, often followed by a signature

oai_token-act-pair · gpt-3.5-turbo

expressions of strong emotions or reactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 estranged

-0.84

hement

-0.81

 honoured

-0.81

 orchestr

-0.80

 ousted

-0.78

 extrad

-0.78

ĸļ

-0.76

 marching

-0.76

 sacked

-0.74

 internationally

-0.74

POSITIVE LOGITS

Conclusion

1.14

CONCLUS

0.97

Example

0.97

Trivia

0.94

Discussion

0.93

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

0.93

Posted

0.92

{{

0.92

References

0.91

Pros

0.91

Activations Density 1.502%

sentences that end with a positive affirmation or appreciation, often followed by a signature

expressions of strong emotions or reactions

No Comments

No Known Activations