INDEX

Explanations

instances of duplication or duplicate-related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to duplication and redundancy

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

Hay

-0.77

Hay

-0.75

 comfort

-0.73

soon

-0.70

 Strait

-0.67

 liberation

-0.66

Tan

-0.64

hill

-0.64

 passers

-0.64

inian

-0.63

POSITIVE LOGITS

 dupl

3.75

 duplicate

3.34

 duplication

3.03

 Dupl

2.57

dup

1.44

 redundant

1.42

Du

1.33

 copies

1.28

 copy

1.25

 clones

1.24

Activations Density 0.030%

instances of duplication or duplicate-related terms

terms related to duplication and redundancy

No Comments

No Known Activations