INDEX

Explanations

adjectives or nouns related to controversial, suspected, or challenging situations

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

lang

-0.91

wright

-0.90

ertodd

-0.89

someone

-0.86

ometers

-0.84

speak

-0.83

rollers

-0.82

rea

-0.82

aido

-0.81

 Feet

-0.80

POSITIVE LOGITS

 inability

1.06

 absence

1.05

 contribution

1.00

 injunction

0.98

 tendency

0.96

 animosity

0.95

 influx

0.95

 arrival

0.95

 threat

0.94

 lack

0.94

Activations Density 2.793%

adjectives or nouns related to controversial, suspected, or challenging situations

No Comments

No Known Activations