INDEX

Explanations

links or connections between different elements or concepts

oai_token-act-pair · gpt-3.5-turbo

references to "link" in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

doms

-0.74

assi

-0.71

boarding

-0.69

 Apocalypse

-0.67

hetti

-0.64

apor

-0.64

sei

-0.63

secut

-0.62

oln

-0.61

 wastes

-0.61

POSITIVE LOGITS

 link

3.87

 links

2.82

link

2.45

 linking

2.22

 Link

2.10

 LINK

2.08

 linkage

1.94

 Links

1.91

links

1.90

Link

1.83

Activations Density 0.013%

links or connections between different elements or concepts

references to "link" in various contexts

No Comments

No Known Activations