INDEX

Explanations

objects or specific items being described in detail in a narrative

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

Politics

-0.89

Financial

-0.80

Daily

-0.78

ivals

-0.78

Online

-0.78

 Markets

-0.78

Ops

-0.75

 Parties

-0.74

MQ

-0.74

IQ

-0.74

POSITIVE LOGITS

 protr

1.74

 thickness

1.61

 diameter

1.46

 thicker

1.39

 seams

1.37

 seam

1.28

 width

1.27

 curved

1.27

 ridge

1.26

 edges

1.26

Activations Density 2.985%

objects or specific items being described in detail in a narrative

No Comments

No Known Activations