INDEX

Explanations

tokens related to news, politics, and current events

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.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

 incorpor

-0.90

 manif

-0.80

 ropes

-0.78

 levers

-0.76

 synerg

-0.75

 reflex

-0.74

 appropri

-0.74

 vulner

-0.74

 tides

-0.74

 slic

-0.73

POSITIVE LOGITS

ï¸ı

1.47

âĶĢâĶĢ

1.22

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

1.08

âĶĢâĶĢâĶĢâĶĢ

0.94

âĹ

0.89

âĸ¬âĸ¬

0.85

ï¸

0.84

Â°

0.83

×Ķ

0.82

STER

0.82

Activations Density 0.307%

tokens related to news, politics, and current events

No Comments

No Known Activations