INDEX

Explanations

phrases relating to understanding processes and experiences in various contexts

oai_token-act-pair · gpt-4o-mini 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

nesty

-0.72

cott

-0.63

starter

-0.62

ritical

-0.61

)]

-0.60

obin

-0.58

folder

-0.57

 Donation

-0.57

 Canaver

-0.56

>>>>>>>>

-0.56

POSITIVE LOGITS

 workings

0.93

 facets

0.87

 unfolding

0.85

 intric

0.84

 surroundings

0.81

 everyday

0.80

 humankind

0.79

 trenches

0.79

 nuances

0.76

 interconnected

0.76

Activations Density 0.627%

phrases relating to understanding processes and experiences in various contexts

No Comments

No Known Activations