INDEX

Explanations

phrases related to direction or position

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

uala

-0.75

 Built

-0.63

 Chimera

-0.61

USE

-0.61

 Quote

-0.60

lance

-0.60

SourceFile

-0.60

pload

-0.59

Ca

-0.59

/>

-0.59

POSITIVE LOGITS

fetched

0.77

endars

0.77

 distances

0.77

 recess

0.74

outheast

0.71

field

0.71

ricular

0.70

izons

0.69

river

0.69

 than

0.68

Activations Density 0.096%

phrases related to direction or position

No Comments

No Known Activations