INDEX

Explanations

terms related to training, preparedness, and techniques, especially in the context of law enforcement and safety

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

bum

-0.70

hao

-0.64

ported

-0.62

pard

-0.62

 headlines

-0.60

rants

-0.60

adra

-0.60

beck

-0.59

 outl

-0.58

leased

-0.58

POSITIVE LOGITS

 taught

1.11

learn

1.08

 lesson

1.07

 lessons

1.05

 instructors

1.01

 instructor

1.01

Learning

0.99

 teach

0.99

 instructional

0.95

 manuals

0.94

Activations Density 2.839%

terms related to training, preparedness, and techniques, especially in the context of law enforcement and safety

No Comments

No Known Activations