INDEX

Explanations

the word "know" used in various contexts

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

ŃĶ

-0.81

ciating

-0.77

phrine

-0.76

ItemTracker

-0.74

edia

-0.71

otion

-0.70

oulos

-0.68

coni

-0.67

ishable

-0.65

attery

-0.65

POSITIVE LOGITS

ledge

1.06

 beforehand

0.99

how

0.98

lege

0.97

ledged

0.86

 firsthand

0.84

 ABOUT

0.82

 exactly

0.80

 whats

0.77

ingly

0.76

Activations Density 0.050%

the word "know" used in various contexts

No Comments

No Known Activations