INDEX

Explanations

scientific and technical information related to various topics, including mathematics, physics, software development, and invention patents

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

die

-0.81

elight

-0.74

rican

-0.73

owship

-0.70

ega

-0.68

fires

-0.67

luster

-0.66

watch

-0.65

igham

-0.65

eligible

-0.65

POSITIVE LOGITS

 equations

1.02

 outline

1.00

 explanations

0.96

 workings

0.95

WHY

0.94

 diagrams

0.91

 Explan

0.90

 explanation

0.90

 outlines

0.90

ĸļ

0.90

Activations Density 3.258%

scientific and technical information related to various topics, including mathematics, physics, software development, and invention patents

No Comments

No Known Activations