INDEX

Explanations

references or citations in an academic context

oai_token-act-pair · gpt-3.5-turbo

reference indicators or citation links

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 mindset

-0.66

ufact

-0.63

 dominating

-0.61

 scrim

-0.61

 sophistication

-0.59

 rece

-0.59

 hous

-0.58

 ploy

-0.58

 ready

-0.57

 autom

-0.57

POSITIVE LOGITS

References

1.29

âĨĳ

1.27

External

1.12

Notes

1.07

Further

1.04

ibliography

1.01

Sources

0.92

Books

0.92

Wik

0.89

Reference

0.88

Activations Density 0.069%

references or citations in an academic context

reference indicators or citation links

No Comments

No Known Activations