INDEX

Explanations

terms related to political and socioeconomic discussions, including topics like surveillance, job creation, public health, and cultural heritage

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

 Borders

-0.78

 Ital

-0.71

 Arabian

-0.70

 Kern

-0.70

Ell

-0.70

 Norris

-0.70

 Burr

-0.69

 Cotton

-0.67

 Cree

-0.67

 Salvador

-0.67

POSITIVE LOGITS

related

1.63

based

1.60

driven

1.55

themed

1.48

loving

1.47

induced

1.44

laden

1.43

oriented

1.42

specific

1.40

derived

1.39

Activations Density 0.441%

terms related to political and socioeconomic discussions, including topics like surveillance, job creation, public health, and cultural heritage

No Comments

No Known Activations