INDEX

Explanations

the term "Res" or related words like "resolve" and "resolution"

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

glers

-0.74

 Holmes

-0.65

 Dynamics

-0.65

 Cruiser

-0.63

tips

-0.62

thing

-0.61

matically

-0.60

 Welsh

-0.60

ratulations

-0.60

 Rove

-0.60

POSITIVE LOGITS

ervation

1.28

olved

1.22

olutions

1.22

idential

1.20

olve

1.19

erves

1.18

earcher

1.18

istance

1.17

erved

1.16

olver

1.14

Activations Density 0.677%

the term "Res" or related words like "resolve" and "resolution"

No Comments

No Known Activations