INDEX

Explanations

specific terms related to a product or project being described, such as reward tiers, development details, and specific project names

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

aws

-0.69

hig

-0.68

licts

-0.68

adows

-0.66

ãĤ¦

-0.66

itiz

-0.66

gas

-0.66

ĥ

-0.65

akable

-0.65

mble

-0.65

POSITIVE LOGITS

 Labs

0.70

 Nova

0.63

's

0.63

 himself

0.62

 Programming

0.61

âĪ

0.61

 fundamentals

0.58

 exec

0.58

 itself

0.58

0.56

Activations Density 0.222%

specific terms related to a product or project being described, such as reward tiers, development details, and specific project names

No Comments

No Known Activations