INDEX

Explanations

words related to medical conditions and physical health, specifically focusing on issues related to nutrition, deficiencies, diseases, and treatment

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

Dash

-0.67

Sah

-0.63

 Loft

-0.63

IJ

-0.63

WOOD

-0.62

 Kali

-0.61

FIX

-0.61

WOR

-0.61

BUR

-0.61

wine

-0.60

POSITIVE LOGITS

cular

0.86

acters

0.82

phy

0.80

illance

0.80

inery

0.79

ication

0.77

emic

0.75

nant

0.75

acies

0.75

anism

0.74

Activations Density 0.126%

words related to medical conditions and physical health, specifically focusing on issues related to nutrition, deficiencies, diseases, and treatment

No Comments

No Known Activations