INDEX

Explanations

phrases related to procedures, processes, and detailed explanations

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aily

-0.78

eday

-0.74

enko

-0.70

iously

-0.68

EVA

-0.67

ieu

-0.66

ilion

-0.66

anth

-0.66

ogether

-0.65

ingly

-0.65

POSITIVE LOGITS

 speaking

0.95

 unchanged

0.83

 consists

0.79

 consisted

0.76

 comprised

0.73

 Marketable

0.73

 consisting

0.73

 consist

0.72

 housed

0.72

 priced

0.71

Activations Density 1.312%

phrases related to procedures, processes, and detailed explanations

No Comments

No Known Activations