INDEX

Explanations

terms related to optimization and efficiency

oai_token-act-pair · gpt-3.5-turbo

terms related to optimization processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soever

-0.87

-0.83

ĸļ

-0.83

FACE

-0.77

hold

-0.74

pole

-0.72

hips

-0.72

ere

-0.71

rough

-0.71

lain

-0.68

POSITIVE LOGITS

 optimization

1.04

 optimize

1.04

 optimizing

0.88

 optim

0.87

 Optim

0.87

bably

0.82

 optimizations

0.80

 allocation

0.78

 gearing

0.72

 optimized

0.71

Activations Density 0.023%

terms related to optimization and efficiency

terms related to optimization processes

No Comments

No Known Activations