INDEX

Explanations

information related to the German language

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

Sharp

-0.81

Moore

-0.77

Matthew

-0.75

Spoiler

-0.72

Cold

-0.72

Bloom

-0.72

Luck

-0.71

RAW

-0.70

Background

-0.70

Daddy

-0.69

POSITIVE LOGITS

qui

1.09

los

1.05

mi

1.02

si

1.00

ni

0.99

que

0.99

 Ã©

0.98

 alle

0.97

est

0.95

Ã

0.95

Activations Density 2.321%

information related to the German language

No Comments

No Known Activations