INDEX

Explanations

mentions and discussions about possibilities or potential scenarios

oai_token-act-pair · gpt-3.5-turbo

references to potential outcomes or scenarios

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

hen

-0.69

ney

-0.65

IT

-0.59

ow

-0.59

ey

-0.59

bell

-0.58

 Chapters

-0.57

 summar

-0.57

Aid

-0.57

ai

-0.56

POSITIVE LOGITS

 possibility

3.78

 possibilities

2.21

ossibility

2.01

 likelihood

1.87

 chance

1.68

 possible

1.66

 impossibility

1.59

 probability

1.56

 prospect

1.53

 feasibility

1.49

Activations Density 0.008%

mentions and discussions about possibilities or potential scenarios

references to potential outcomes or scenarios

No Comments

No Known Activations