INDEX

Explanations

references to specific topics/issues within discussions or debates

oai_token-act-pair · gpt-3.5-turbo

references to specific topics and issues in discussions or debates

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

icons

-0.82

nces

-0.81

asters

-0.75

ressor

-0.71

trained

-0.70

izont

-0.69

renheit

-0.69

icone

-0.68

rough

-0.68

agues

-0.67

POSITIVE LOGITS

 topic

1.99

 issue

1.99

 question

1.87

 matter

1.80

 subject

1.76

 issues

1.32

 topics

1.30

 Question

1.29

matter

1.29

 matters

1.28

Activations Density 0.390%

references to specific topics/issues within discussions or debates

references to specific topics and issues in discussions or debates

No Comments

No Known Activations