INDEX

Explanations

phrases related to events or conditions happening after a specific trigger or event

oai_token-act-pair · gpt-3.5-turbo

references to post-event or post-condition scenarios

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 chants

-0.83

 outnumbered

-0.83

 propositions

-0.72

 Wanted

-0.72

 laure

-0.71

 latitude

-0.71

 gauge

-0.70

 disadvantaged

-0.69

 Leader

-0.69

 sorely

-0.69

POSITIVE LOGITS

mortem

1.70

traumatic

1.61

secondary

1.46

election

1.45

processing

1.45

mort

1.35

doctoral

1.35

graduate

1.32

production

1.32

truth

1.30

Activations Density 0.018%

phrases related to events or conditions happening after a specific trigger or event

references to post-event or post-condition scenarios

No Comments

No Known Activations