INDEX

Explanations

words related to exploration, especially in a technical or adventurous context

oai_token-act-pair · gpt-3.5-turbo

references to the concept of exploration across 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

Haf

-0.73

loo

-0.71

sup

-0.68

signed

-0.65

Kev

-0.64

estic

-0.64

 Emanuel

-0.64

lam

-0.64

lam

-0.63

 Serve

-0.63

POSITIVE LOGITS

 exploration

3.65

 Exploration

2.66

 explor

2.40

 explorers

1.97

 exploring

1.95

 explore

1.78

 explorer

1.78

 experimentation

1.51

 discoveries

1.48

 discovery

1.47

Activations Density 0.018%

words related to exploration, especially in a technical or adventurous context

references to the concept of exploration across various contexts

No Comments

No Known Activations