INDEX

Explanations

information and instructions on a specific topic or task

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

 scorp

-0.64

 manner

-0.64

riel

-0.63

 adamant

-0.63

 interim

-0.63

 chast

-0.62

ber

-0.62

aign

-0.62

 climax

-0.61

 elusive

-0.60

POSITIVE LOGITS

 Nope

1.14

Try

1.11

 Wouldn

1.09

 Well

1.08

 Congratulations

1.06

Yep

1.04

 Consider

1.04

 Turns

1.03

 Check

1.02

 Probably

1.01

Activations Density 0.065%

information and instructions on a specific topic or task

No Comments

No Known Activations