INDEX

Explanations

phrases or words related to the main or biggest points in a discussion or context

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

renheit

-0.80

each

-0.75

FIL

-0.71

late

-0.69

PET

-0.68

ipel

-0.67

rub

-0.67

livion

-0.67

nery

-0.67

azard

-0.66

POSITIVE LOGITS

 reason

1.33

 drawback

1.32

 caveat

1.32

 downside

1.31

 thing

1.26

 takeaway

1.26

 difference

1.26

 distinguishing

1.24

 question

1.24

 lesson

1.14

Activations Density 0.147%

phrases or words related to the main or biggest points in a discussion or context

No Comments

No Known Activations