INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

references to locations or entities, particularly significant places and their associated contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

 charact

-0.74

 encount

-0.70

 describ

-0.70

 destro

-0.69

ÃĥÃĤÃĥÃĤ

-0.68

 Turtles

-0.67

jri

-0.67

Franc

-0.66

 unden

-0.66

abe

-0.65

POSITIVE LOGITS

 Miscellaneous

0.78

 Diary

0.76

 Conclusion

0.75

ãģ®å®

0.73

 Discussion

0.71

 Composite

0.69

å·

0.69

 Purchase

0.68

Reviewer

0.68

 Recommend

0.67

Activations Density 0.101%

proper nouns

references to locations or entities, particularly significant places and their associated contexts

No Comments

No Known Activations