INDEX

Explanations

phrases related to writing and jotting down ideas or thoughts

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

ï¿½ï¿½ï¿½

-0.74

terday

-0.71

ï¿½ï¿½

-0.69

forge

-0.69

anwhile

-0.66

antry

-0.63

externalActionCode

-0.60

rouse

-0.58

staking

-0.58

 therap

-0.58

POSITIVE LOGITS

 excess

0.89

 chunks

0.77

 roots

0.76

 scraps

0.75

 copies

0.74

 tons

0.74

 unwanted

0.71

 submissions

0.69

 quantities

0.69

 tonnes

0.69

Activations Density 0.131%

phrases related to writing and jotting down ideas or thoughts

No Comments

No Known Activations