INDEX

Explanations

phrases related to consequences and decision-making

oai_token-act-pair · gpt-3.5-turbo

statements or phrases indicating consequences or outcomes

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 scheduled

-0.85

 awaited

-0.83

 planned

-0.83

 officially

-0.77

 pione

-0.76

 earthqu

-0.75

ported

-0.75

 rumored

-0.74

ascus

-0.73

iliated

-0.73

POSITIVE LOGITS

 Therefore

1.43

 Hence

1.38

 Especially

1.37

 Unless

1.36

 Sometimes

1.35

 Anything

1.32

 Whereas

1.27

 Consequently

1.24

If

1.24

 Otherwise

1.24

Activations Density 0.508%

phrases related to consequences and decision-making

statements or phrases indicating consequences or outcomes

No Comments

No Known Activations