INDEX

Explanations

news-related terms and headlines

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ļ½

-0.64

umenthal

-0.63

heit

-0.62

 homebrew

-0.61

eele

-0.60

kinson

-0.59

onson

-0.58

wagon

-0.57

é¾įå¥ĳå£«

-0.56

metics

-0.56

POSITIVE LOGITS

 MUST

0.96

BUT

0.65

eals

0.65

 Ùħ

0.61

could

0.60

quickShipAvailable

0.60

AND

0.59

 Replay

0.59

must

0.57

 WATCH

0.56

Activations Density 0.009%

news-related terms and headlines

No Comments

No Known Activations