INDEX

Explanations

technology-related terms, specifically related to software development and machine learning

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

 Concern

-0.78

soType

-0.73

 Beir

-0.66

 disappoint

-0.65

Ake

-0.64

 depl

-0.62

 Horn

-0.60

pite

-0.60

 failings

-0.60

Viol

-0.60

POSITIVE LOGITS

 seamlessly

1.35

 customizable

1.21

 seamless

1.09

 instantaneous

1.07

 automatically

1.06

 instantly

1.06

 unlimited

1.03

 effortlessly

1.02

 customize

1.00

 access

0.99

Activations Density 3.044%

technology-related terms, specifically related to software development and machine learning

No Comments

No Known Activations