INDEX

Explanations

Roman numerals followed by a physical medicine condition.

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

hem

-0.63

 doubtful

-0.62

 upside

-0.62

 blame

-0.61

 readiness

-0.61

 trouble

-0.61

 transitioning

-0.60

 appropriate

-0.60

 limp

-0.60

 isol

-0.60

POSITIVE LOGITS

Sys

0.82

acus

0.79

oto

0.76

 Tayyip

0.74

ãĤ¨ãĥ«

0.73

Corp

0.72

ãĤ¿

0.71

Serv

0.71

asonic

0.70

ott

0.70

Activations Density 0.188%

Roman numerals followed by a physical medicine condition.

No Comments

No Known Activations