INDEX

Explanations

references to specific people, events, or entertainment works

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

 disag

-0.88

ulic

-0.81

Statistics

-0.81

 consolidate

-0.81

 centralized

-0.80

 administr

-0.79

 consolid

-0.79

osate

-0.77

ritis

-0.76

cling

-0.75

POSITIVE LOGITS

 actress

1.43

 actor

1.40

 films

1.40

 movies

1.39

 sitcom

1.39

 actors

1.36

 Conan

1.32

 actresses

1.32

 movie

1.32

HBO

1.31

Activations Density 3.444%

references to specific people, events, or entertainment works

No Comments

No Known Activations