INDEX

Explanations

terms related to optimization and efficiency

oai_token-act-pair · gpt-3.5-turbo

terms related to optimization and efficiency

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

ouse

-0.76

 Commonwealth

-0.74

 sworn

-0.74

usp

-0.74

 Membership

-0.73

 Rape

-0.73

hold

-0.72

 detain

-0.72

ewitness

-0.71

Iss

-0.71

POSITIVE LOGITS

 optimization

3.17

 optim

3.06

 optimize

2.85

 optimizing

2.85

 optimized

2.73

optim

2.66

 Optim

2.65

 optimizations

2.61

 optimal

2.23

Opt

1.93

Activations Density 0.033%

terms related to optimization and efficiency

terms related to optimization and efficiency

No Comments

No Known Activations