INDEX

Explanations

phrases related to opinions, decision-making, and conclusions

oai_token-act-pair · gpt-3.5-turbo

concepts related to claims, discussions, and decision-making processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 hemor

-0.73

RET

-0.63

BUT

-0.60

iren

-0.58

SEE

-0.58

Advertisements

-0.56

ãĤ»

-0.53

 behold

-0.53

 pestic

-0.53

 Inher

-0.52

POSITIVE LOGITS

 requires

1.63

 entails

1.50

 involves

1.47

 helps

1.32

 reduces

1.28

 implies

1.28

 ensures

1.27

 allows

1.26

isn

1.24

 constitutes

1.24

Activations Density 0.629%

phrases related to opinions, decision-making, and conclusions

concepts related to claims, discussions, and decision-making processes

No Comments

No Known Activations